Objectives Integrating vocational rehabilitation and mental healthcare has shown effects on vocational outcomes during sick leave with common mental disorders. In a previous paper, we showed that a Danish integrated healthcare and vocational rehabilitation intervention (INT) had a surprisingly negative impact on vocational outcomes compared to service as usual (SAU) at 6-and 12-month follow-up. That was also the case with a mental healthcare intervention (MHC) tested in the same study. This article reports the 24-month follow-up results of that same study. Method A randomized, parallel-group, three-arm, multi-centre superiority trial was conducted to test the effectiveness of INT and MHC compared to SAU. Results In total, 631 persons were randomized. Contrary to our hypothesis, SAU showed faster return to work than both INT [hazard rate (HR) 1.39, P=0.0027] and MHC (HR 1.30, P=0.013) at 24-month follow-up. Overall, no differences were observed regarding mental health and functional level. Compared to SAU, we observed some health benefits of MHC, but not INT, at 6-month follow-up but not thereafter, and lower rates of employment at all follow-ups. Since implementation problems might explain the results of INT, we cannot conclude that INT is no better that SAU. The MHC intervention was implemented with good fidelity and did not improve return to work. Conclusion This trial does not support the hypothesis that INT lead to faster return to work. However, implementation failure may explain the negative results.