Penile plethysmography is considered a promising procedure for the assessment of sexual arousal and orientation. However, the discovery that some individuals are capable of invalidating the results of the test has raised concerns about its limitations. The purpose of this study was to investigate the effectiveness of a signal detection task in eliminating faking. Additionally, the effect of an incentive on the ability to fake on penile plethysmography was evaluated. A total of 32 heterosexual subjects participated in the study and 11 homosexual subjects were involved in the replication phase. Overall, subjects evidenced a significant ability to suppress erections to preferred sexual stimuli. The incentive did not affect performance. The signal detection procedure appeared to interfere with erection under standard instructions but did not significantly reduce faking ability. As predicted, there was a tendency for recognition scores to be lower under faking conditions. Intensive cognitive processing may facilitate (rather than eliminate) faking and enhance recall of test stimuli.