Smoking habits in adolescence have a strong impact on smoking habits in adulthood. Nine of ten adult smokers started smoking before the age of eighteen. Thus, efforts to prevent the use of tobacco among minors are of great interest. In 1997 a minimum-age law of 18 years for purchase of tobacco was introduced in Sweden. The purpose of the age limit was to reduce the availability of tobacco to young people and hence ultimately to reduce the consumption of tobacco.

The overall purpose of the work reported in the present thesis was:

• to study whether the Swedish minimum-age law for the purchase of tobacco has affected young people’s possibilities of buying tobacco
• to study how compliance with the minimum-age law for the purchase of tobacco can be improved
• to develop methods for following compliance with the minimum-age law for the purchase of tobacco.

The empirical basis for the work was three different types of data:

1. Test purchases of tobacco in 1996, 1999, 2002 and 2005 in Värmland, Västernorrland and Malmö. In total there were forty-eight test purchasers that made 3,150 test purchases.
2. Questionnaire surveys on tobacco habits and attitudes that were carried out in 1996, 2000 and 2005 among students in grades 7 and 9 of the compulsory school and in 1996 and 2000 in grade 2 of the upper-secondary school. All data was collected in the same three regions as the test purchase studies. The questionnaire was completed by a total of about 57,000 students.
3. Structured telephone interviews in 2005 with 28 key people in tobacco prevention work in the three regions.

In conclusion, the thesis shows that:

• the minimum-age law introduced in 1997 for the purchase of tobacco products has limited adolescents’ opportunities for purchasing tobacco. Compliance with the minimum-age law is, however, still incomplete
• the most important factor in whether test purchasers were able to buy tobacco was whether there was an age control (as the law prescribes)
• compliance with the minimum-age law differs among the three regions studied. It is reasonable to assume that the regional differences can be related partly to regional differences in the effectiveness of the measures taken to ensure compliance with the law
• the proportion of daily smokers among students in grade 9 in the compulsory school declined from 1996 to 2005 in Värmland and Västernorrland but there was no change in Malmö
• the method used for test purchase of tobacco and further developed and adapted for Swedish conditions, functioned well for following compliance with the law.