AIM: Examine negative side-effects on self-rated health and well-being of a community-based intervention in a fishing community in the Norwegian Arctic. METHOD: A cohort study with quasi-experimental design with one intervention community and three control communities from the same area. There was a baseline screening (1987), three years intervention (1988-1991) and a rescreening (1993). Of the invited (aged 20-62 years in 1987), 668 (64%) males and 656 (72%) females met at both screenings. The main outcome variables were self-rated health, mental health variables, and preoccupation with health. Those answering the relevant questions at both screenings were analysed. High-risk and low-risk males were analysed separately. There were too few females in the high-risk group to do separate analyses. RESULTS: The main findings among the low-risk group were: Males in Batsfjord did not differ from males in the control communities. Among females, there was a reduction in the proportion depressed of 20.4% points in Batsfjord and 9.6% in the control communities (p adjusted=0.060). There was also a decrease in the proportion lonely of 19.2% points and 5.8% respectively (p adjusted=0.005). In the male high-risk group there was a decrease in the proportion lonely by 14.7% points in Batsfjord and an increase by 5.9% in the control communities (p adjusted=0.004). CONCLUSION: We have not found that an intervention programme using local empowerment, mass strategies, and intervention on high-risk cases has serious negative side effects on self-rated health, well-being or preoccupation with health.