AbstractBackground: High prevalence of minor depression and low well-being constitutes both individual and societal burden. The Internet provides a promising platform for treatment delivery.Objective and Method: A randomized controlled trial was conducted to determine the efficacy of an internet-based self-help intervention Bedre Hverdag (BH) for mild-to-moderate depression and promotion of well-being. Long-term effects (i.e.6 months) of BH were assessed in comparison to a waiting-list control group.Results: Of the 206 eligible participants, 112 were randomized to the experimental group and 94 to the waiting-list control group. Data from 34 participants in the experimental group and 47 from the control group were subjected to statistical analyses. Individuals assigned to BH reported significant increases in positive affect (PA) at 1 month follow-up, as measured by Positive Affect Schedual (PANAS); z = - 2.49, p = .011, r = .42 , which were significantly different from levels of PA at the same time period in the control group, U = 589, z = -2.01 , p = .044 , r = .2. Increases in PA were no longer significant at 2 and 6 month follow-ups. No statistically significant reductions in negative affect (NA) as measured by the Negative Affect Schedual (PANAS) and depressive symptoms as measured by the Center for Epidemiologic Studies Depression scale (CES-D) were observed in the experimental group as compared to the control group. Treatment x Educational attainment interaction explained 5.3% of variance in depressive symptoms, and 10.7% of variance in NA at 1 month follow-up, above and beyond the main effects. Individuals with lowest level of educational attainment reported largest reductions in both CES-D scores and NA.Conclusion: BH produced significant increases in PA at 1 month post-treatment. No significant reductions were observed in either depressive symptoms or NA. Educational attainment moderated the intervention's effects on depressive symptoms and NA.