Diagnosis of drug exposure is strongly supported by analysis of hair samples. In the province of Brescia, Italy, for regranting driving license to drug addicts or occasional abusers, a control programme was adopted including analysis of illicit drugs in two hair segments (0-3 and 3-6 cm) and in urine. From January 1998 to April 1999, upon request of the Local Medical Commission, 697 hair samples were tested in our laboratory. One hundred and eighty subjects resulted positive in hair for one or two of the controlled drug classes (73.3% for cocaine, 10% for opiates, 16.7% for both). Positive subjects were classified by residence, age, sex and license category. Seventy-two subjects were called back after 6-12 months and submitted to a second hair and urine analysis: in 34 cases the result of the first analysis was confirmed (19 negatives, 15 positives for one or both drug classes). Another 37 cases tested positive at the first control and negative at the second, suggesting the hypothesis that a strict control may have a significant deterrent function. The high percentage of negative results at the second control may be explained by the prevalence of cocaine users in the examined population. Our results allow us to conclude that the strict application of control rules lead to a decrease of social risk behaviours.