OBJECTIVES: This study aimed at describing the frequency of full, partial, and no return to work after long-term sick leave and at ascertaining the influence of psychosocial work conditions, work ability and health, reported before the onset of sick leave, on full and partial return to work. METHODS: Altogether 853 public-sector employees in Sweden, mainly women, with at least one sick leave lasting >/=28 days, were studied. The outcome was the level of sick leave 2 years after the sick leave began. Potential predictors were self-rated health, work ability, and psychosocial work conditions assessed by questionnaire before the sick leave. Odds ratios (OR) and 95% confidence intervals (95% CI) were calculated by multinomial regression analyses. RESULTS: Altogether 41% of the participants went directly from full sick leave to full return to work; 21% had periods of partial return to work, but, at the 2-year follow-up, were fully back to work; 15% had partial return to work; and 23% were still not working. A relaxed work situation, a combination of low demands and high decision latitude, increased the odds for full (OR 2.72, 95% CI 1.60-4.62) and partial (OR 2.42, 95% CI 1.21-4.85) return to work. Negative consequences of organizational changes were associated with decreased odds for full return to work (OR 0.54, 95% CI 0.38-0.77). Good self-rated health and work ability were associated with full return to work. CONCLUSIONS: Partial return to work often precedes full return to work, but also operates as a long-term solution for remaining occupationally active. Promoting relatively low demands and high decision latitude at work may support both full and partial return to work after long sick leaves.