This study investigated the effects of terminating low dose levels of caffeine (100 mg/day) in 7 normal humans. Substitution of placebo capsules for caffeine capsules occurred under double-blind conditions while subjects rated various dimensions of their mood and behavior. In the first phase of the study, substitution of placebo for 12 consecutive days resulted in an orderly withdrawal syndrome in 4 subjects which peaked on days 1 or 2 and progressively decreased toward prewithdrawal levels over about 1 week. Data from the remaining three subjects provided no evidence of withdrawal. In the second phase of the study, the generality of the withdrawal effect was examined by repeatedly substituting placebo for 100 mg/day of caffeine for 1-day periods separated by an average of 9 days. Despite differences within and across subjects with respect to the presence, nature and magnitude of symptoms, each of the seven subjects demonstrated a statistically significant withdrawal effect. Although the phenomenon of caffeine withdrawal has been described previously, the present report documents that the incidence of caffeine withdrawal is higher (100% of subjects), the daily dose level at which withdrawal occurs is lower (roughly equivalent to the amount of caffeine in a single cup of strong brewed coffee or 3 cans of caffeinated soft drink) and the range of symptoms experienced is broader (including headache, fatigue and other dysphoric mood changes, muscle pain/stiffness, flu-like feelings, nausea/vomiting and craving for caffeine) than heretofore recognized.