Research on general deterrence does not seem to indicate that stiffer penalties deter. On the contrary, increased penalties may result in fewer prosecutions and convictions, thereby reducing the certainty of punishment -- and thus indirectly reducing its deterrent efficacy. Moreover, long terms of imprisonment do not seem to be effective in deterring the individual offender. Their main justification is thus based upon their incapacitative effect. This is a costly policy, however, since it is difficult to identify those relatively few sex offenders who will continue to be dangerous to the public after release. And even if identification were possible, their detention would have to be inordinately long if success is to be ensured. A penalty scale based on just deserts does not require uniformly heavy penalties, but rather a differentiation between more and less serious forms of sexual aggression. Finally, although the rehabilitation ideal has been tarnished of late, serious consideration should be given to experimental programs for treatment of sex offenders. (Author abstract modified)