Participants with a criminal history, excluding DWI arrests, were more than twice as likely as those with no criminal history to be rearrested for a DWI within a 24-month period. The average DWI recidivism rates for the treatment completers of 14 percent was considerably lower than the 34 percent for the dropouts. For those with a prior criminal record, the higher DWI recidivism rate proved consistent across treatment dropout and treatment-completed subgroups. The treatment-completed subgroup showed a significantly lower recidivism rate than the dropouts for individuals without a criminal history. The number of times the participants drank and drove was reduced significantly across time, and this effect was related to criminal history. The no-criminal history subgroup demonstrated significant decreases across all time periods, and the criminal history subgroup declined to near zero. Study findings suggest that consideration of DWI subgroups, specifically criminal history, may prove important when assessing the effectiveness of intervention and treatment strategies. 7 footnotes, 1 figure, and 1 table