This study evaluated the effects of a community alcohol prevention program on the frequency of alcohol service to young adults at licensed premises in Stockholm, Sweden. We used a pretest (1996)-posttests (1998 and 2001) design with intervention and control areas. The multicomponent intervention combines training of serving staff in responsible beverage service, policy initiatives, and enforcement of existing alcohol regulations. Adolescents 18 years old (the legal drinking age on licensed premises in Sweden), but younger looking according to an expert panel, visited licensed premises in pairs, where each adolescent ordered a beer. At baseline in 1996, the adolescents made 600 attempts to order. At follow-up in 1998, the number of attempts to order was 252, and at the second follow-up in 2001, the adolescents made 238 attempts. We found no statistically significant differences between the intervention and control areas. Overall, the frequency of alcohol service to adolescents on licensed premises in these areas of Stockholm decreased significantly over time, from 45 to 41 and to 32%, in 1996, 1998, and 2001, respectively. The decrease in alcohol service in 2001 was statistically significant compared to the baseline in 1996. One explanation for this improvement could be more effective enforcement of existing alcohol laws in both the intervention and control areas. We also found that licensed premises that used doormen to screen potential customers were less likely to sell to minors.