Recent revelations of sexual misconduct by sports coaches have challenged long-held beliefs in the educational value of sport, yet there is very little knowledge about the dynamics of sexual exploitation in sport upon which to base improvements in the practice of sports coaching or teaching. Earlier inductive research by Brackenridge (1996, 1997a, 1997b) in Britain established a set of hypothesised risk factors for sexual abuse in sport which have subsequently been reinforced by the results of survey research on elite athletes in Canada (Kirby and Greaves 1996). However, risk analysis for sexual abuse in sport has not yet been framed within a temporal or developmental sequence, nor sufficiently differentiated between elite and recreational levels of sport, or between coach-initiated and peer-initiated abuse. This paper reports selected findings from a Dutch qualitative study (Cense 1997) of 14 athletes who have survived sexual abuse in sport. The aim of the study was to identify risk factors that influence sexual abuse and harassment and to analyse which risks might be diminished through a prevention policy implemented by sport organisations. The Dutch study reinforces the earlier risk factor analyses but extends them by putting forward a preliminary temporal model of risk in sport that integrates offender behaviour with athlete and situational factors. On the basis of this model, suggestions are made to assist early diagnosis and prevention of sexual harassment and abuse by authority figures in sport.