A Sex Difference in the
Speciﬁcity of Sexual Arousal
Meredith L. Chivers,
and J. Michael Bailey
Northwestern University and
Centre for Addiction and Mental Health, Toronto, Ontario, Canada
ABSTRACT––Sexual arousal is category-specific in men;
heterosexual men are more aroused by female than by
male sexual stimuli, whereas homosexual men show the
opposite pattern. There is reason to believe that female
sexual arousal is organized differently. We assessed geni-
tal and subjective sexual arousal to male and female sex-
ual stimuli in women, men, and postoperative male-to-
female transsexuals. In contrast to men, women showed
little category speciﬁcity on either the genital or the sub-
jective measure. Both heterosexual and homosexual
women experienced strong genital arousal to both male
and female sexual stimuli. Transsexuals showed a cate-
gory-specific pattern, demonstrating that category speci-
ﬁcity can be detected in the neovagina using a
photoplethysmographic measure of female genital sexual
arousal. In a second study, we showed that our results for
females are unlikely to be explained by ascertainment bi-
ases. These ﬁndings suggest that sexual arousal patterns
play fundamentally different roles in male and female
Male sexual arousal is category-specific; men show their greatest
sexual arousal to the categories of people with whom they prefer
to have sex. With respect to sexual orientation, heterosexual men
experience much higher genital and subjective arousal to women
than to men, whereas homosexual men show an opposite pattern
(Freund, 1963). Category speciﬁcity is sufﬁciently reliable for
forensic practitioners to use genital sexual arousal patterns to
assess sexual preferences among men who are strongly moti-
vated to conceal their preferences: Examples include pedophiles
(e.g., Quinsey & Lalumie
`re, 2001), in whom greatest sexual
arousal occurs to sexual stimuli depicting prepubescent chil-
dren. Moreover, sexual arousal patterns appear to be an impor-
tant source of information for men as they formulate their sexual
identities in adolescence (Bell, Weinberg, & Hammersmith,
1981; Savin-Williams & Diamond, 2000).
Several lines of evidence suggest that women’s sexual pref-
erences may not be as strongly related to sexual arousal patterns
as men’s preferences are. First, sexual arousal appears to be a
less important signal of nonheterosexual orientation in women
than in men (Savin-Williams & Diamond, 2000). Second, fe-
male sexuality seems generally to be more ﬂexible than male
sexuality, with greater intra-individual variation in preferences,
behavior, attitudes, and responsiveness to cultural inﬂuences
(Baumeister, 2000). Baumeister argued that this partly reﬂects
a weaker female sex drive. Greater ﬂexibility in female sexual
preferences may also be reﬂected in a less specific pattern of
The most direct evidence that female sexual arousal is less
category-specific than male sexual arousal comes from a study
by Laan, Sonderman, and Janssen (1996). They measured the
subjective and genital arousal patterns of self-identiﬁed lesbian
and heterosexual women who viewed ﬁlms depicting male-fe-
male and female-female sex. No effect for self-identiﬁcation (as
lesbian or heterosexual) was observed for either genital or
subjective arousal: Both lesbians and heterosexual women ex-
perienced their highest genital and subjective arousal to male-
Although the ﬁndings of Laan et al. (1996) are intriguing,
their interpretation is not straightforward for three reasons.
First, the sexual stimuli did not include a pure male stimulus.
The literature on male sexual arousal suggests that the most
effective contrast is between arousal to an intense, purely fe-
male stimulus (typically a ﬁlm of female-female sex) and
arousal to an intense, purely male stimulus (typically a ﬁlm of
male-male sex; Mavissakalian, Blanchard, Abel, & Barlow,
1975; Sakheim, Barlow, Beck, & Abrahamson, 1985). Sexual
stimuli depicting male-female couples fail to elicit significant
differences in sexual arousal patterns because they contain both
men and women. Second, it is unclear whether the sexual ori-
entation (as opposed to the sexual identity; see Mustanski,
Chivers, & Bailey, 2002) of the self-identiﬁed lesbians was fully
homosexual or bisexual. For example, Rust (1992) found that
Address correspondence to Meredith L. Chivers, Clinical Sexology
Services, Centre for Addiction and Mental Health, 250 College St.,
Toronto, Ontario, Canada M5T 1R8; e-mail: meredith_chivers@
736 Volume 15—Number 11Copyright r2004 American Psychological Society
30% of women who identiﬁed themselves as lesbian reported
some sexual attraction to men. A more specific measure of
sexual orientation, one assessing degree of attraction to both
men and women, would have been preferable for classifying
participants. Third, although women’s genital sexual arousal can
be objectively measured by photoplethysmography (e.g., Janssen,
2002; Laan & Everaerd, 1995), research supporting construct
validity is less abundant for photoplethysmographic measures
than for male phallometric measures (Janssen, 2002). Before
concluding that women do not have a category-specific sexual
arousal pattern, it would be desirable to demonstrate that vaginal
measures are, in principle, capable of detecting such a pattern.
In the present study, we examined whether female sexual
arousal is category-specific. We assessed sexual arousal pat-
terns to male versus female sexual stimuli in women, men, and
an identiﬁed subset of women (postoperative male-to-female
transsexuals). Including male participants allowed us to com-
pare male and female arousal patterns and to demonstrate that
our stimuli were capable of eliciting a category-specific pattern
of sexual arousal in men. Including male-to-female transsexuals
allowed us to determine whether differences in arousal patterns
between men and women merely reﬂect differences in the way
that genital arousal is measured in men and women, or are due
to a true sexual dimorphism.
We recruited heterosexual and homosexual men and women via
advertisements in an alternative urban newspaper (Chicago
Reader) and in publications from the gay and lesbian community.
Participants were 69 men and 52 women. Mean ages were 32.1
(SD 56.0) and 26.4 (SD 56.5) years for the male and female
samples, respectively. Transsexual participants were identiﬁed
as a subset of women participants. The mean age of the trans-
sexual sample (n511) was 42.9 (SD 510.5) years. All par-
ticipants were offered ﬁnancial compensation for participation.
We used the Kinsey Sexual Fantasy Scale (Kinsey, Pomeroy,
Martin, & Gebhard, 1953) to assess the sexual preferences of all
participants. Only individuals indicating exclusive or nearly
exclusive sexual feelings for either women or men during
adulthood were included in the analyses. It was possible to
identify male-to-female transsexuals with preferences for male
or female sexual partners because there are two distinct sub-
types of male-to-female transsexuals: One is exclusively sexu-
ally attracted to men, whereas the other is primarily sexually
attracted to women (Blanchard, 1989, 1992).
Measures and Materials
Audiovisual Stimuli. Films of sexual stimuli elicit greater
genital and subjective arousal than do either slides or audio-
tapes (Heiman, 1977; McConaghy, 1999). We chose ﬁlms as our
stimuli to ensure that participants experienced substantial
genital arousal responses and thus avoid ﬂoor effects. We used
ﬁlms of male-male and female-female sex because men’s sexual
orientation is most reliably assessed by comparing penile re-
sponses to these stimuli (Mavissakalian et al., 1975; Sakheim
et al., 1985). Although male-female stimuli are less discrimi-
nating, we included them to assess whether heterosexual par-
ticipants’ arousal to purely male or female stimuli might be
diminished by the fact that purely male or female stimuli depict
homosexual acts, which are stigmatized in mainstream Ameri-
can society. If so, arousal to male-female stimuli should exceed
that to homosexual stimuli.
The sexual stimuli consisted of six 2-min ﬁlms with sound.
Content varied by the sex of the actors (male or female) and the
type of sexual activity depicted (oral or penetrative sex). Each
participant saw ﬁlms featuring female-female oral sex, female-
female penetration (with a strap-on dildo), male-female cunni-
lingus, male-female penetration (penile-vaginal), male-male
fellatio, and male-male penetration (penile-anal). Thus, sex of
actors and type of sexual activity were independent. Each
participant viewed one of two exemplars from each stimulus
category, with stimuli presented in a random order. A neutral
stimulus, depicting landscapes or fauna, was also included, to
provide a nonsexual comparison for genital and subjective re-
sponses to sexual stimuli. An 11-min adaptation ﬁlm (depicting
sexually neutral scenes accompanied by relaxing music) was
used to assess baseline arousal.
Psychophysiological Assessment. All psychophysiological data
were continuously recorded and digitized using an MP100WS data-
acquisition unit (BIOPAC Systems Inc., Goleta, California) and
AcqKnowledge software, Version 3.2.7 (BIOPAC Systems Inc.).
Male genital arousal was assessed with penile plethysmography
(Janssen, 2002) using a mercury-in-rubber strain gauge to measure
changes in the circumference of the penis as erection developed.
The signal was low-pass ﬁltered (to 0.5 Hz) and digitized (40 Hz).
The gauge was calibrated over six 5-mm steps between sessions
(Janssen, 2002). The penile plethysmograph signal was trans-
formed into millimeters of circumference change from baseline.
Women’s and male-to-female transsexuals’ genital arousal was
assessed via change in vaginal pulse amplitude (VPA), a mea-
sure of vaginal vasocongestion specific to sexual response (Laan,
Everaerd, & Evers, 1995), using a vaginal photoplethysmograph
(Sintchak & Geer, 1975). The VPA signal was band-pass ﬁltered
(0.5 Hz to 10 Hz) and digitized (40 Hz). VPA was measured as
peak-to-trough amplitude for each vaginal pulse.
The neovagina of a postoperative male-to-female transsexual
consists of a lined neovaginal cavity within the perineum. The
lining is typically constructed from a penile skin ﬂap, although it
is sometimes constructed from scrotal skin, a segment of intes-
tinal tissue, or a skin graft from another location (Karim, Hage, &
Mulder, 1996). The neovagina is surrounded, in whole or in part,
Volume 15—Number 11 737
M.L. Chivers et al.
by the highly vascular tissue of the male pelvis, including
periurethral erectile tissue that is likely homologous to the
erectile tissue surrounding the urethra in biological females.
Although a prior study found detectable blood ﬂow in the neo-
vaginal lining (Schroder & Carroll, 1999), this was the ﬁrst at-
tempt to determine whether blood ﬂow increases with psycho-
logical sexual stimulation.
Subjective Arousal. Subjective sexual arousal was assessed
continuously via self-report using a lever moving through a 1801
arc; 01represented no subjective sexual arousal, and 1801rep-
resented the subjective sexual arousal associated with orgasm.
The signal was low-pass ﬁltered (to 0.5 Hz) and digitized (40 Hz).
The lever signal was transformed into percentage deﬂection.
Participants were assessed individually in a dimly lit, private
room, seated in a comfortable recliner with a television monitor
5 ft away. Participants received instruction on how to use
the genital gauge, and they ﬁtted the gauge themselves. They
watched the adaptation ﬁlm and then the experimental stimuli
(sexual and neutral), separated by return-to-baseline intervals.
Participants completed distraction tasks during interstimulus
intervals and, after assessment of sexual arousal, completed
questionnaires assessing their sexual orientation, sexual expe-
rience, masturbation frequency, and orgasmic capacity.
Both genital and subjective arousal measures were averaged,
separately and within stimulus category, yielding mean genital
and subjective arousal for responses to female-female, male-
female, and male-male sexual stimuli. Mean scores were
standardized within subjects (i.e., ipsatized) because within-
subjects standardization appears to eliminate the effects of idi-
osyncratic variation in responsiveness (Harris, Rice, Quinsey,
Chaplin, & Earls, 1992). An index of arousal to male relative to
female sexual stimuli, the male-female contrast, was computed,
separately for genital and subjective arousal, by subtracting
arousal to female-female stimuli from arousal to male-male
stimuli; positive scores indicated greater arousal to male stimuli,
and negative scores indicated greater arousal to female stimuli.
Genital and subjective arousal to females was computed as the
difference between arousal to female-female stimuli and to the
neutral stimulus; genital and subjective arousal to males and to
male-female stimuli was computed analogously.
Because not all participants produce a discernible genital
response to sexual stimuli, for all three samples we used an
inclusion criterion of a minimum difference of 0.5 standard
deviations between maximum genital arousal to either male or
female stimuli and to the neutral stimulus. Additionally, on the
basis of recommendations of other researchers (Seto et al.,
2001), we excluded men whose maximum response to either
male or female stimuli did not exceed their response to the
neutral stimulus by at least 2 mm. These criteria excluded 23 of
69 men, 9 of 52 women, and none of the 11 transsexuals. The
difference in rates of exclusion was significant (Fisher’s exact
test, p5.02), probably owing to differences in the sensitivity of
the penile plethysmograph and the photoplethysmograph. The
exclusion rate in our male sample (approximately 1 in 3) is
typical of phallometric assessments using circumferential measure-
ment (e.g., Kuban, Barbaree, & Blanchard, 1999). Inclusion of
nonresponders did not substantially affect the significance or
direction of results.
The relation between sexual orientation and patterns of genital
sexual arousal to male versus female stimuli, by sample, is
presented in Figure 1. Table 1 presents, for each sample, the
correlations among dichotomous self-reported sexual prefer-
ence, genital arousal (male-female contrast), and subjective
arousal (male-female contrast). In general, the relation between
self-reported preference and sexual arousal pattern was much
weaker for women than for men and transsexuals, whose results
were similar. For example, all transsexuals and nearly all men
had stronger genital arousal to their preferred sex than to their
nonpreferred sex, but 37% of women did not. The correlation
between self-reported preference and genital arousal was sig-
nificantly lower for women than for men (z55.0, p<.001) and
for transsexuals (z58.9, p<.001). The analogous correlations
for subjective arousal were also lower for women than for the
other samples, although the difference was significant only for
men (z55.3, p<.001; for transsexuals, z51.1, p5.27).
Table 1 also shows that the association between genital and
subjective arousal was lower for women than for men (z53.9,
p<.001), replicating a commonly observed sex difference
(Laan & Everaerd, 1995).
A plausible concern is that pure sexual stimuli depict ho-
mosexual interactions, which may be unusual or repugnant
stimuli for some heterosexual people. Thus, heterosexual peo-
ples’ responses to pure male or female sexual stimuli might
underestimate their arousability to members of the opposite sex.
If this is true, then heterosexual participants might be expected
to respond more strongly to stimuli depicting male-female acts
than to pure opposite-sex stimuli depicting same-sex acts. The
genital arousal data for men and women, however, do not show
this pattern (Fig. 2). Heterosexual men were more aroused,
genitally and subjectively, by ﬁlms depicting female-female sex
acts than by ﬁlms depicting male-female sex acts, t(21) 53.0,
p<.01, d51.08, and t(21) 52.9, p<.01, d51.16, re-
spectively. Similarly, heterosexual women (see Fig. 2) were
slightly, but not significantly, more genitally aroused by ﬁlms
depicting male-male sex acts than by ﬁlms depicting male-
female sex acts, t(22) 50.9, p5.4, d50.24. With respect to
subjective arousal, heterosexual women did report a strong
preference for male-female stimuli: Their subjective sexual
738 Volume 15—Number 11
Sex Difference in Sexual Arousal
arousal to the male-female stimuli was approximately three
times greater than their subjective arousal to the male-male
stimuli, t(22) 54.8, p<.001, d51.62.
Because many individuals, particularly women, are reluctant to
be assessed genitally (Wolchik, Spencer, & Iris, 1983), it was
impossible to recruit a random sample for Study 1. Therefore, it
is plausible that the nonspecific arousal pattern observed in
women was inﬂuenced by ascertainment bias. Several studies
have examined ascertainment bias in female sexual psycho-
physiological research (e.g., Morokoff, 1986; Wolchik, Braver,
& Jensen, 1985; Wolchik et al., 1983). Compared with refusers,
volunteers for sexual arousal research masturbate more often,
have more experience with sexual materials, and have had more
sex partners. The fact that volunteers differ from refusers does
not necessarily mean, however, that volunteers do not represent
the general population with respect to their patterns of sexual
arousal: That is a separate empirical question. If differences
between volunteers and refusers are relevant to arousal pat-
terns, then variables related to cooperation should be correlated
with indices of sexual arousal. In Study 2, we examined whether
the results for females in Study 1 were plausibly due to ascer-
tainment bias. We invited a different sample of women to par-
ticipate in a subsequent study of genital sexual arousal and
examined the differences between women who refused to par-
ticipate and those who agreed to participate on several variables
identiﬁed in previous research as potentially relevant. We then
correlated these variables with sexual arousal patterns among
the women who participated in the sexual arousal study. We also
sought to replicate the nonspecific pattern of arousal observed
in heterosexual women in Study 1.
We asked 232 women from undergraduate psychology classes to
attend an information session describing a sexual arousal study,
and 104 attended. After hearing the details of the study, these
women completed an anonymous questionnaire assessing their
sexual experiences and their interest in participating in the
arousal study. Of the 104 women who attended the information
session, 57 stated they were interested and 47 stated they were
Fig. 1. Mean genital arousal to male versus female stimuli for men, women, and male-to-female transsexuals as a
function of self-reported preferred sex (Study 1). The center horizontal lines of the distributions represent the re-
spective means, and the upper and lower lines represent the boundaries of the 95% conﬁdence intervals. Points
represent individual participants. Units are within-subjects standard deviations.
Within-Group Correlations Among Genital Arousal and
Subjective Arousal Contrasts and Self-Reported Sexual
Preference, Study 1
Subjective arousal contrast .48 ––
Prefer men versus women .26 .42
Subjective arousal contrast .88 ––
Prefer men versus women .88 .92
Subjective arousal contrast .67 ––
Prefer men versus women .96 .70
Note. Ranges represent 95% conﬁdence intervals.
Volume 15—Number 11 739
M.L. Chivers et al.
not interested in participating in the study. Of the 57 women
who indicated an interest during the information session, 29
participated in the sexual arousal study. (Six of these women
were not included in the analyses because they either did not
meet a minimum response criterion or reported a nonhetero-
sexual sexual preference.) The 29 participants in the arousal
study completed sexual-experience questionnaires twice, both in
the information session and during the arousal study, but par-
ticipant anonymity precluded our knowing which questionnaires
from the information session came from women who eventually
participated in the arousal study. All other methods and proce-
dures were identical to those described for females in Study 1.
The 23 heterosexual students whose data were included in the
sexual arousal study were significantly younger (M520.2,
SD 51.0 vs. M526.4, SD 56.5, respectively) than the het-
erosexual women from Study 1, F(1, 22) 520.7, p<.05, and
had significantly fewer male sexual partners (M52.5, SD 5
1.9 vs. M57.5, SD 57.9, respectively), F(1, 22) 58.5,
p<.05. All the women reported having sexual feelings almost
exclusively toward men; their mean Kinsey Sexual Fantasy
score was 0.1 (SD 50.1).
The heterosexual students showed the same nonspecific
arousal pattern (Fig. 3) that we found in the community sample
of women. Although all Study 2 participants preferred male sex
partners, their genital arousal was 19% higher to female-female
stimuli than to male-male stimuli, t(22) 51.9, p5.07,
d50.64, and their subjective arousal to female-female stimuli
was twice their subjective arousal to male-male stimuli,
t(22) 54.9, p<.001, d51.54. Similar to the heterosexual
female participants from Study 1, the students were subjectively
more aroused to the ﬁlms depicting heterosexual acts than to
the male-male stimuli, t(22) 58.4, p<.001, d53.31, and
the female-female stimuli, t(22) 57.1, p<.001, d52.37. The
strong subjective preferences for male-female stimuli did
not translate into genital differences that were as strong.
Genital arousal to the male-female stimuli was 22% higher than
genital arousal to the male-male stimuli, t(22) 52.1, p5.04,
Fig. 2. Mean genital (upper graphs) and subjective (lower graphs) arousal to male-male, female-female, and male-female sexual
stimuli (minus responses to the neutral stimulus), for men, women, and male-to-female transsexuals, as a function of self-reported
preferred sex (Study 1). Units are within-subjects standard deviations. Error bars show standard errors of the mean.
740 Volume 15—Number 11
Sex Difference in Sexual Arousal
d50.79, and 3% higher than genital arousal to the female-
female stimuli, t(22) 50.3, p5.76, d50.11.
Several sexual-history variables were related to willingness to
participate in a study requiring measurement of genital arousal
(Table 2). To explore whether differences between students who
participated in the arousal study and those who refused to
participate inﬂuenced arousal patterns, we examined correla-
tions between the variables related to willingness to participate
and the arousal contrasts (Table 3). The correlations were
generally small and revealed no clear pattern. Three significant
correlations showed that higher frequency of orgasm during
masturbation was associated with higher genital arousal to
male-male stimuli, higher subjective arousal to female-female
stimuli, and more subjective arousal to female-female relative
to male-male stimuli. Thus, there is no convincing evidence that
volunteer biases led to a misleading picture of female sexual
DISCUSSION AND CONCLUSIONS
Our ﬁndings suggest that women have a nonspecific pattern of
sexual arousal that is quite different from men’s category-spe-
cific pattern. Men and postoperative male-to-female transsex-
uals preferring men showed substantially higher subjective and
genital responses to male-male than to female-female stimuli,
and men and transsexuals preferring women showed the oppo-
site pattern. In contrast, women’s subjective and genital re-
sponses were only modestly related to their preferred category:
Heterosexual and lesbian women experienced genital and
subjective arousal to both male-male and female-female stim-
uli. Our ﬁndings suggest that this result is not plausibly at-
tributable to volunteer biases. Variables that distinguished
female volunteers from females who refused to participate were
unrelated to response patterns.
The difference between men’s and women’s sexual arousal
patterns is unlikely to be due to measurement artifacts because
women and transsexuals had different genital arousal patterns
despite being measured by the same apparatus, and because
transsexuals and men had similar genital arousal patterns despite
being measured by a different apparatus. Other evidence for a
fundamental difference between women’s and men’s sexual
arousal patterns comes from their patterns of subjective sexual
arousal. Regardless of sexual orientation, women reported more
arousal to female-female than to male-male stimuli, on average,
although this difference was not significant for heterosexual
women. In contrast, men and transsexuals reported being more
aroused by sexual stimuli corresponding to their preferred gender.
Although our results suggest that women have a nonspecific
pattern of arousal to sexual stimuli, they do not imply that women’s
sexual orientation is inherently bisexual. For example, despite
their capacity to become sexually aroused by both male and fe-
male sexual stimuli, women do not have higher rates of same-sex
sexual activity than men (Laumann, Gagnon, Michael, & Mi-
chaels, 1994). The large majority of women in contemporary
Western societies have sex exclusively with men (Bajos
et al., 1995; Laumann et al., 1994). Similarly, the large majority of
women in these societies report much higher attraction to men
than to women (Bailey, Dunne, & Martin, 2000; Laumann et al.,
1994). A self-identiﬁed heterosexualwoman would be mistaken to
question her sexual identity because she became aroused
watching female-female erotica; most heterosexual women expe-
rience such arousal. A self-identiﬁed heterosexual man who ex-
perienced substantial arousal to male-male erotica, however,
would be statistically justiﬁed in reconsidering his sexual identity.
Sex differences in the development (Bell et al., 1981; Dia-
mond, 2000, 2003; Savin-Williams & Diamond, 2000) and
expression (Baumeister, 2000) of same-sex attraction support
our contention that the relation between sexual arousal and
sexual orientation differs fundamentally between women and
men. In the context of past research, our results suggest that
patterns of sexual arousal to men versus women do not constrain
women’s sexual behavior, feelings, or identity to nearly the
degree that they constrain men’s.
Our results cannot directly address whether the sex differ-
ence in category speciﬁcity of sexual arousal is innate or
learned. Our ﬁnding that male-to-female transsexuals show a
Fig. 3. Heterosexual female students’ genital and subjective sexual
arousal to male-male, female-female, and male-female sexual stimuli
(minus responses to the neutral stimulus) in Study 2. Units are within-
subjects standard deviations. Error bars show standard errors of the
Volume 15—Number 11 741
M.L. Chivers et al.
male-typical pattern, however, helps to rule out some possible
explanations. Women’s nonspecific pattern might not be fully
explained by their lack of visible genitalia because transsexuals
show a category-specific pattern despite a similar lack. Trans-
sexuals reject the male gender role into which they were so-
cialized yet continue to show the category-specific pattern of
arousal that is characteristic of their genetic sex. Moreover, the
biological or social factors that cause some transsexuals to be
the most feminine of males (Bailey, 2003) do not affect their
male-typical pattern of sexual arousal. This ﬁnding is consistent
with other evidence that psychosexual differentiation is multi-
dimensional (e.g., Bailey, 2003; Bailey, Gaulin, Agyei, & Gla-
due, 1994; Goy, Bercovitch, & McBrair, 1988).
One potential methodological limitation of our study con-
cerns the nature of the sexual stimuli we used. The ‘‘pure’’ sexual
stimuli depicted homosexual acts, which some heterosexual
Sexual-Experience Variables Significantly Discriminating Among Women Expressing Varying Levels
of Interest in Study Participation
Number of male sexual
M(SD) 0.5 (0.7) 2.2 (2.3) 2.6 (2.5) 14.2 (<.001)
Range 0–2 0–12 0–11
M(SD) 1.7 (3.2) 5.8 (5.4) 6.8 (5.4) 13.1 (<.001)
Range 0–15 0–20 0–20
Frequency of orgasm during
M(SD) 4.4 (2.5) 5.7 (1.9) 5.4 (2.1) 3.1 (.051)
Range 1–7 1–7 1–7
Preferred frequency of sex
M(SD) 3.0 (1.3) 3.8 (1.1) 3.9 (1.0) 6.0 (.003)
Range 1–6 1–6 1–6
Frequency of erotica use
M(SD) 1.9 (1.1) 2.9 (1.2) 3.3 (1.1) 15.6 (<.001)
Range 1–5 1–6 1–5
For this variable, ns521, 48, and 25, respectively. These values represent women who reported masturbating at least
once in a 30-day period. The scale for this variable was as follows: 1 5never; 2 5<20%; 3 520–39%; 4 540–59%;
5560–79%; 6 580–99%; 7 5100%.
The scale for this variable was as follows: 1 5<1 time/week; 2 51 time/week; 3 52
times/week; 4 53–5 times/week; 5 5daily; 6 5> 1 time/day.
The scale for this variable was as follows: 1 5never;
25watched once; 3 5watched twice; 4 51 time/year; 5 51 time/month; 6 51 time/day.
Correlations Among Potentially Relevant Sexual-Experience Variables and Arousal Contrasts for Study 2 Participants
Genital arousal contrast Subjective arousal contrast
M-M vs. N F-F vs. N M-F vs. N M-M vs. F-F M-M vs. N F-F vs. N M-F vs. N M-M vs. F-F
Number of male sexual partners
(penile-vaginal intercourse) –.31 .18 .12 .31 .13 .07 .27 .07
Masturbation frequency, per month .23 .31 .26 .35 .01 –.08 .24 .05
Frequency of orgasm during
.18 .03 .42 .19 .59
Preferred frequency of sex .04 –.04 .11 .05 .11 .10 .16 .04
Frequency of erotica use .26 .02 .03 .14 .35 .05 .05 .25
Note. M-M 5male-male stimuli; F-F 5female-female stimuli; M-F 5male-female stimuli; N5neutral stimulus.
742 Volume 15—Number 11
Sex Difference in Sexual Arousal
people may ﬁnd upsetting or even offensive. Although the
comparison of arousal to homosexual versus female-male stimuli
suggested that aversion to homosexual acts did not greatly di-
minish genital arousal, it would be desirable to replicate our study
using sexual stimuli that do not include homosexual depictions.
Films of individuals engaged in sexual acts, such as masturbation,
would avoid this interpretative difﬁculty while retaining the
sexual intensity necessary to elicit sufﬁcient sexual arousal.
A second limitation is our assumption that participants did
not consciously manipulate their genital responses. Past re-
search has demonstrated that some men can control their genital
arousal when they are motivated to do so (Adams, Motsinger,
McAnulty, & Moore, 1992; Freund, 1963). Therefore, men’s
category-specific sexual arousal may be due to conscious in-
hibition of responses to nonpreferred sexual stimuli. Although
the design of our study cannot rule out this possibility, we think
it is unlikely for several reasons. First, in contexts where men
are highly motivated to manipulate their sexual arousal, such
as during phallometric assessment for pedophilia, category-
specific sexual arousal usually occurs (Blanchard, Klassen,
Dickey, Kuban, & Blak, 2001). Second, although heterosexual
male participants might be motivated to suppress sexual arousal
to male stimuli because homosexuality is stigmatized, gay men
would not be similarly motivated to suppress arousal to female
stimuli; yet gay men’s sexual arousal was also category-specific.
Third, research on the conscious manipulation of genital sexual
arousal has shown that men are able to reduce, but not increase,
the magnitude of their erections. Neither heterosexual nor ho-
mosexual men are able to increase erections to nonpreferred
sexual stimuli, even when motivated to do so (Adams et al., 1992).
This suggests that under typical conditions, men are not sup-
pressing their genital responses to nonpreferred stimuli; if they
were, they would also increase genital arousal to these stimuli, by
ceasing their suppression efforts, when motivated to do so.
The sex difference reported here has important implications
for future conceptualizations of women’s sexuality. Sexual
arousal, especially genital sexual arousal, likely plays a much
smaller role in women’s sexual-orientation development than it
does in men’s. Female sexuality, in general, may be more mo-
tivated by extrinsic factors, such as the desire to initiate or
maintain a romantic relationship, than by intrinsic factors, such
as genital sexual arousal (Baumeister, Catanese, & Vohs, 2001).
This basic sex difference in the role of sexual arousal processes
highlights the need to use distinct models when investigating
the development and expression of female and male sexuality.
Acknowledgments––National Institute of Child and Human
Development Grant R03-HD41770-01 and a grant from the
American Psychological Foundation funded this research. We
lenstein, M. Lalumie
`re, A. Lawrence, S. LeVay, J. Panksepp, and M.
Seto for helpful comments, and we thank the study participants.
Adams, H.E., Motsinger, P., McAnulty, R.D., & Moore, A.L. (1992).
Voluntary control of penile tumescence among homosexual and
heterosexual subjects. Archives of Sexual Behavior,21, 17–31.
Bailey, J.M. (2003). The man who would be queen. Washington, DC:
Joseph Henry Press.
Bailey, J.M., Dunne, M.P., & Martin, N.G. (2000). Genetic and envi-
ronmental inﬂuences on sexual orientation and its correlates in an
Australian twin sample. Journal of Personality and Social Psy-
Bailey, J.M., Gaulin, S., Agyei, Y., & Gladue, B.A. (1994). Effects of
gender and sexual orientation on evolutionarily relevant aspects
of human mating psychology. Journal of Personality and Social
Bajos, N., Wadsworth, J., Ducot, B., Johnson, A.M., Le Pont, F., Wel-
lings, K., Spira, A., & Field, J. (1995). Sexual behaviour and HIV
epidemiology: Comparative analysis in France and Britain. The
ACSF Group. AIDS,9, 735–743.
Baumeister, R.F. (2000). Gender differences in erotic plasticity: The
female sex-drive as socially ﬂexible and responsive. Psychologi-
cal Bulletin,126, 347–374.
Baumeister, R.F., Catanese, K.R., & Vohs, K.D. (2001). Is there a
gender difference in strength of sex drive? Theoretical views,
conceptual distinctions, and a review of relevant evidence. Per-
sonality and Social Psychology Review,5, 242–273.
Bell, A.P., Weinberg, M.S., & Hammersmith, S.K. (1981). Sexual
preference: Its development in men and women. Bloomington, IN:
Alfred C. Kinsey Institute of Sex Research.
Blanchard, R. (1989). The concept of autogynephilia and the typology
of male gender dysphoria. Journal of Nervous and Mental Disease,
Blanchard, R. (1992). Nonmonotonic relation of autogynephilia and
heterosexual attraction. Journal of Abnormal Psychology,101,
Blanchard, R., Klassen, P., Dickey, R., Kuban, M.E., & Blak, T. (2001).
Sensitivity and speciﬁcity of the phallometric test for pedophilia
in nonadmitting sex offenders. Psychological Assessment,13,
Diamond, L.M. (2000). Sexual identity, attractions, and behavior
among young sexual-minority women over a 2-year period. De-
velopmental Psychology,36, 241–250.
Diamond, L.M. (2003). What does sexual orientation orient? A
biobehavioral model distinguishing romantic love and sexual
desire. Psychological Review,110, 173–192.
Freund, K. (1963). A laboratory method for diagnosing predominance
of homo- or hetero-erotic interest in the male. Behaviour Research
and Therapy,1, 85–93.
Goy, R.W., Bercovitch, F.B., & McBrair, M.C. (1988). Behavioral mas-
culinization is independent of genital masculinization in prena-
tally androgenized female rhesus macaques. Hormones and Be-
Harris, G.T., Rice, M.E., Quinsey, V.L., Chaplin, T.C., & Earls, C.
(1992). Maximizing the discriminant validity of phallometric
assessment data. Psychological Assessment,4, 502–511.
Heiman, J.R. (1977). A psychophysiological exploration of sexual
arousal patterns in females and males. Psychophysiology,14,
Janssen, E. (2002). Psychophysiological measures of sexual response.
In M.W. Wiederman & B.E. Whitley (Eds.), Handbook for
Volume 15—Number 11 743
M.L. Chivers et al.
conducting research on human sexuality (pp. 139–171). Mahwah,
Karim, R.B., Hage, J.J., & Mulder, J.W. (1996). Neovaginoplasty in
male transsexuals: Review of surgical techniques and recom-
mendations regarding eligibility. Annals of Plastic Surgery,37,
Kinsey, A.C., Pomeroy, W.B., Martin, C.E., & Gebhard, P.H. (1953).
Sexual behavior in the human female. Philadelphia: Saunders.
Kuban, M., Barbaree, H.E., & Blanchard, R. (1999). A comparison of
volume and circumference phallometry: Response magnitude and
method agreement. Archives of Sexual Behavior,28, 345–359.
Laan, E., & Everaerd, W. (1995). Determinants of female sexual
arousal: Psychophysiological theory and data. Annual Review of
Sex Research,6, 32–76.
Laan, E., Everaerd, W., & Evers, A. (1995). Assessment of female
sexual arousal: Response speciﬁcity and construct validity. Psy-
Laan, E., Sonderman, M., & Janssen, E. (1996, June). Straight and
lesbian women’s sexual responses to straight and lesbian erotica: No
sexual orientation effects. Poster presented at the annual meeting
of the International Academy of Sex Research, Rotterdam, The
Laumann, E.O., Gagnon, J.H., Michael, R.T., & Michaels, S. (1994).
The social organization of sexuality: Sexual practices in the United
States. Chicago: University of Chicago Press.
Mavissakalian, M., Blanchard, E.B., Abel, G.C., & Barlow, D.H.
(1975). Responses to complex erotic stimuli in homosexual and
heterosexual males. British Journal of Psychiatry,126, 252–257.
McConaghy, N. (1999). Unresolved issues in scientific sexology. Ar-
chives of Sexual Behavior,28, 285–318.
Morokoff, P.J. (1986). Volunteer bias in the psychophysiological study
of female sexuality. Journal of Sex Research,22, 35–51.
Mustanski, B., Chivers, M.L., & Bailey, J.M. (2002). A review and
critique of the evidence for a biological basis of human sexual
orientation. Annual Review of Sex Research,13, 89–140.
Quinsey, V.L., & Lalumie
`re, M.L. (2001). Assessment of sexual offenders
against children. Thousand Oaks, CA: Sage.
Rust, P.C. (1992). The politics of sexual identity: Sexual attraction and
behavior among lesbian and bisexual women. Social Problems,39,
Sakheim, D.K., Barlow, D.H., Beck, J.G., & Abrahamson, D.J. (1985).
A comparison of male heterosexual and male homosexual patterns
of sexual arousal. Journal of Sex Research,21, 183–198.
Savin-Williams, R.C., & Diamond, L.M. (2000). Sexual identity tra-
jectories among sexual-minority youths: Gender comparisons.
Archives of Sexual Behavior,29, 607–627.
Schroder, M., & Carroll, R.A. (1999). New women: Sexological out-
comes of male-to-female gender reassignment surgery. Journal of
Sex Education and Therapy,24, 137–146.
Seto, M.C., Adkerson, D.L., Hindman, J., Jensen, S.H., Peters, J.M., &
Peterson, K.D. (2001). Practice standards and guidelines for
members of the Association for the Treatment of Sexual Abusers (3rd
ed.). Beaverton, OR: Association for the Treatment of Sexual
Sintchak, G., & Geer, J.H. (1975). A vaginal plethysmograph system.
Wolchik, S.A., Braver, S.L., & Jensen, K. (1985). Volunteer bias in
erotica research: Effects of intrusiveness of measure and sexual
background. Archives of Sexual Behavior,14, 93–107.
Wolchik, S.A., Spencer, S.L., & Iris, I.S. (1983). Volunteer bias in re-
search employing vaginal measures of sexual arousal. Archives of
Sexual Behavior,12, 399–408.
(RECEIVED 11/11/02; REVISION ACCEPTED 4/23/04)
744 Volume 15—Number 11
Sex Difference in Sexual Arousal