Objective: To evaluate the effectiveness of a population-based multi-factorial lifestyle intervention on long-term changes in dietary habits compared to a non-intervention control group. Methods: The study was a randomized controlled lifestyle intervention study, Inter99 (1999-2006), Copenhagen, Denmark, using a high-risk strategy. Participants in the intervention group (n = 6 091) had at baseline a medical health-examination and a face-to-face lifestyle counselling. Individuals at high risk of ischemic heart disease were repeatedly offered both individual and group-based counselling. The control group (n = 3 324) was followed by questionnaires. Dietary habits were measured by a validated 48-item food frequency questionnaire and changes were analyzed by multilevel analyses. Results: At the 5-year follow-up the intervention group compared to the control group had significantly increased their intake of vegetables (men: net-change: 23 g/week; p = 0.04; women: net-change: 27 g/week; p = 0.005) and decreased the intake of highly saturated fats used on bread and for cooking (men: OR = 0.59 (0.41-0.86); women: OR = 0.42 (0.30-0.59)). Significant effects on fruit and fish intake were found at the 3-year follow-up but the effect attenuated at the 5-year follow-up. Conclusion: A population-based multi-factorial lifestyle intervention promoted significant greater beneficial long-term dietary changes compared to the control group, especially the intake of vegetables and saturated fat was improved.