A review of previous research on criminal deterrence notes that few findings support the notion that individuals punished for a crime are deterred from subsequent offenses, or that specific deterrence is a function of severity of punishment. The specific deterrent effect of judicial penalties on a sample of 1,021 drivers convicted of a breathalyzer offense in New South Wales in 1972 was measured by determining the rate of reconviction for the same offense within a 2-year period and by gauging the perceived severity of penalty through weighing personal characteristics of the offenders and the seriousness of their offenses. In cases of 'average' seriousness, nearly eight times as many people who received a low severity penalty were reconvicted as those who received a heavy penalty. Findings for the small number of cases in the least serious and most severe categories were too equivocal to determine the impact of penalties on offenders. Separate analyses conducted to measure the effect of imprisonment and license disqualification upon offenders showed a high reconviction rate for motorists previously sentenced to prison and a high percentage of repeat offenders among drivers whose driving licenses had been revoked for 1 year. However, much lower recidivism rates were evident among disqualified drivers with revoked licenses for periods of less than 1 year. In conclusion, the question of deterrence is closely related to the nature of the drinking driver in that heavy penalties correlate with lower recidivism rates among some classes of offenders. Further analysis is required to separate the effect of offender characteristics from the effect of penalties. Four tables, 1 figure, and 20 references are provided.