The effectiveness of a health education programme for prevention of alcohol abuse and tobacco smoking among adolescents was assessed by an intervention study with a non-randomized control group. The study was carried out among 7th-9th grade school students in a Local Health Unit of the Brescia province, North Italy. Students were enrolled in either the intervention or the control group, based on the participation of their teachers to the educational programme. A total of 428 students were involved in the programme representing the intervention group, and 658 students not involved in the programme formed the control group. The health education programme included the prevention of alcohol abuse in the first school year (1989-90) and the prevention of tobacco smoking in the second year (1990-91). Each part of the programme included a series of lessons taken by the teachers for about 20 hours. Alcohol use and smoking habits, knowledge about, and attitudes toward alcohol and tobacco use of the students have been assessed by anonymous questionnaires. The questionnaires have been administered during school time at the beginning and the end of each school year, giving a total of 6 questionnaires, from Q1 to Q6. All the questionnaires included the same questions on alcohol and tobacco use, knowledge and attitudes. Of the 1086 students who attended the 7th grade in the year 1989-90, 428 were enrolled as intervention group and 658 as control group. As regards alcohol, the following findings were observed: 1. no difference was found in the percentages of students drinking at least one glass of wine or beer daily between the 2 groups at Q6; 2. no substantial change in the percentage of students aware of the health risks of alcohol abuse from Q1 to Q6 was observed. As regards tobacco smoking, the results of the study showed: 1. no difference between the 2 groups was found at Q6 as regards the proportion of students who smoked one or more cigarettes a month; 2. the percentages of students planning to smoke in the future were similar in the 2 groups in Q6; 3.the percentage of students with a high score on the health consequences of smoking was higher in the intervention than the control group at Q4, but again no difference between the groups was evident at Q6. In conclusion, the health education programme seemed to be unsuccessful in modifying behaviors and attitudes regarding alcohol and tobacco use during the 3-year follow-up.