OBJECTIVES: Previously, we reported that an inpatient multimodal occupational rehabilitation program (I-MORE) was more effective than outpatient Acceptance and Commitment Therapy (O-ACT) in reducing sickness absence and was cost-effective over a 24-month period. Here we present 7-years of follow-up on sick leave and the cost of lost production. METHODS: We randomized individuals aged 18-60, sick-listed due to musculoskeletal or mental health disorders to I-MORE (n = 82) or O-ACT (n = 79). I-MORE, lasting 3.5 weeks, integrated ACT, physical training, and work-related problem-solving. In contrast, O-ACT mainly offered six weekly 2.5 h group sessions of ACT. We measured outcomes using registry data for days on medical benefits and calculated costs of lost production. Our analysis included regression analyses to examine differences in sickness absence days, logistic general estimating equations for repeated events, and generalized linear models to assess differences in costs of lost production. RESULTS: Unadjusted regression analyses showed 80 fewer days of sickness absence in the 7-year follow-up for I-MORE compared to O-ACT (95% CI - 264 to 104), with an adjusted difference of 114 fewer days (95% CI - 298 to 71). The difference in costs of production loss in favour of I-MORE was 27,048 euros per participant (95% CI - 35,009 to 89,104). CONCLUSIONS: I-MORE outperformed O-ACT in reducing sickness absence and production loss costs during seven years of follow-up, but due to a limited sample size the results were unprecise. Considering the potential for substantial societal cost savings from reduced sick leave, there is a need for larger, long-term studies to evaluate return-to-work interventions.