Sexual offenders who had been referred for a forensic evaluation were assessed in terms of their deviant sexual arousal by means of standard penile plethysmographic procedures. The sample was divided into Rapist and Nonrapist groups according to their sexual history and criminal charges. Penile circumference responses (PCR) and self-reported arousal to consenting sex, sexual assault, and physical assault were monitored. The assessments were repeated following subject instructions to suppress their arousal and following the ingestion of alcohol. The overall correlation between PCR and self-reported arousal was .65. An Alcohol-by-Offender Type interaction revealed that the lowering effect of alcohol on PCR occurred only among the Nonrapists. The suppression instruction increased Rapists' PCRs to Rape and Physical Assault presentations relative to consenting sexual narrations. When the data were examined in relation to intelligence, low IQ Rapists displayed greater responses to rape than high IQ Rapists. Under the influence of alcohol, low IQ Rapists displayed greater arousal regardless of the stimulus, while high IQ Rapists showed no change, and Nonrapists responded less than they did without alcohol. Comments concerning the accurate assessment of deviant sexual arousal and the identification of precursors to deviant sexual behaviour are made in view of the current findings.