Background. Several large and well-conducted community interventions have failed to detect an effect on prevalence of smoking. Methods. Two thousand four hundred eight daily smokers in all motivational stages were actively recruited and included in a randomised population-based intervention study in Copenhagen, Denmark. All smokers completed a questionnaire and underwent a health examination and a lifestyle consultation. Daily smokers in the high intensity intervention group were offered assistance to quit in smoking cessation groups. Results. The validated abstinence rate at 1-year follow-up was 16.3% in the high intensity group and 12.7% in the low intensity group compared with a self-reported abstinence rate of 7.3% in the background population. The adjusted odds ratio of abstinence in the high intervention group was significantly higher, OR = 2.2 (1.6-3.0) than in the background population, also in the 'intention-to-treat' analyses, OR = 1.5 (1.1-2.0). Higher socioeconomic status, higher age at onset of daily smoking, and a higher wish to quit were predictors of success. Conclusion. In a population-based setting, using active recruitment and offering assistance to quit, it was possible to include many smokers and to achieve a significantly higher validated abstinence in the high intensity intervention than in the background population, even when using 'intention-to-treat' analyses.