The National Institute of Forensic Toxicology (NIFT) in Oslo receives blood samples from all Norwegian drivers suspected of driving under the influence of alcohol. It is well known that a large proportion of the arrested drunken drivers are repeat offenders. The purposes of this investigation was to find the arrest rates (the percentage of subjects arrested once or more) among drunken drivers followed retrospectively and prospectively during the 11-year period 1984-1994 and the probability of 'abstaining' from becoming a recidivist during the 9 years subsequent to the year of selection. By examining the rearrest rates during the 3 following years for drivers selected in 1986, 1989, 1991 and 1992 we tried to look for major effects due to the change in the Norwegian road traffic act of 1988. Altogether 45% of the selected drunken drivers were arrested two or more times. Totally the '9-year survival rate' (i.e. not being rearrested) was 60% for drivers with blood alcohol concentration (BAC) selected from the interval 0.06-0.09%; 56% from BAC 0.13-0.16% and 51% from 0.26-0.29%. The data were further evaluated with respect to frequency of rearrest during 3 years after selection, and was around 30% in 1986, while it was lower for drivers selected in 1992 (19%). An explanation for the reduction in rearrest rate may be the changes in the road traffic act which took place in 1988.