Using long-term DUI (Driving Under the Influence of Alcohol) arrest recidivism data from a controlled study of DUI intervention effectiveness, interactions among DUI interventions, age, race, education, and alcohol severity were estimated using logit analysis. Data were collected in a 9-year follow-up study of the Mississippi DUI Probation Project. The effects of short-term interventions (alcohol education schools for low alcohol severity offenders and structured group interventions for high alcohol severity offenders) were specified by educational level. Short-term rehabilitation was modestly effective for those with less than 12 years of education, but less effective or detrimental for the more highly educated. The effects of probation were specified by age and education, being more effective for those under 30 years and 55 years or older than for the middle aged group. Probation was most effective for well-educated older (55 +) offenders. An analysis of the under 30 years group also suggested that probation was especially effective for young well-educated Minority offenders.