This article describes (i) gender differences in initial return to work (RTW) and lasting return to work (LRTW); (ii) health characteristics that predicted the observed gender differences in LRTW; and (iii) associations between health characteristics and (L)RTW in men and women separately. A 13-month prospective cohort study was performed with 119 employees (54 women and 65 men) who had reported sick for more than 1 month because of mental or musculoskeletal complaints. These employees were interviewed bimonthly about their work status and health. The health characteristics included were self-rated health, the reason for reporting sick, the presence of at least one long-term disease, early improvement in health and a change in diagnosis. Cox regression analyses on the time to (L)RTW were performed. No gender differences were found regarding RTW. Women, however, did report a longer time to LRTW than men. This was predicted by their reason for reporting sick, the presence of at least one long-term disease, lack of early improvement in health and change in diagnosis. Women who experienced no early improvement and/or whose diagnosis changed during their sickness absence had lower chances of attaining a LRTW than other women. Men who were absent owing to mental complaints were less likely to RTW than men with musculoskeletal complaints. Men who had at least one long-term disease were also less likely to RTW than men who had no long-term diseases. The findings showed a delayed LRTW for women when compared with men, as predicted by health characteristics. The health characteristics, however, that predicted LRTW in women differed from those that predicted LRTW in men. Clearly, the findings support a gender-specific approach to sickness absence guidance. The results also suggest that women are misdiagnosed more often than men. This requires further investigation.