Road traffic collisions caused by drunk driving pose a significant public health problem all over the world. Therefore additional preventive activities against drunk driving should be worked out. The aim of the study was to assess drunk driving in novice drivers after a psychological intervention taking into account also impulsivity, law obedience, and alcohol-related measures. An intervention study was started with 1889 car driver's license attempters during their driving school studies. Subjects were classified as intervention group (n=1083, mean age 23.1 (SD=7.4) years), control group (n=517, mean age 22.8 (SD=7.1) years) and "lost" group (n=289, mean age 23.0 (SD=6.9) years). "Lost" group subjects had been assigned into the intervention group, but they did not participate in the intervention. Subjects of the intervention group participated in a psychological intervention on the dangers of impulsive behavior in traffic. After a three year follow-up period it appeared that in the control group and in the lost group there was a significantly higher proportion of drunk drivers than in the intervention group, 3.3% (n=17), 3.5% (n=10) and 1.5% (n=10) (p=0.026), respectively. Survival analysis confirmed that psychological intervention had a significant impact on drunk driving (p=0.015), and the impact of the intervention was persistent also in the case of higher scores in Mild social deviance. In subjects with higher scores in impulsivity measures and alcohol-related problems the impact of short psychological intervention was not sufficient for preventing drunk driving. It can be concluded that psychological intervention used during the driving school studies is an effective primary prevention activity against drunk driving. However, for drivers with high scores in impulsivity measures and alcohol-related problems, the short psychological intervention is not sufficient in reducing drunk driving behavior.