OBJECTIVE Recent research has emphasized that return to work (RTW) is a dynamic, gradual and often uneven process with a great degree of individual variation. This study aimed to identify RTW trajectories of Swedish employees on sick-leave due to common mental disorders (CMDs). The second aim was to explore which demographic, employment, health-related and work environment characteristics predicted RTW trajectory membership. METHODS Data comes from two 2-armed cluster-randomized controlled trials (RCT) with a 12-month follow-up. A participative problem-solving intervention aimed to reduce sick-leave was compared to care as usual (CAU) involving any kind of work-directed interventions. Participants on sick-leave due to CMDs at baseline (N = 197) formed the study sample. Latent growth mixture modeling and logistic regression were the main analytical approaches. RESULTS Five distinct RTW trajectories of Swedish employees were identified: Early RTW (N = 65), Delayed RTW (N = 50), Late RTW (N = 39), Struggling RTW (N = 21) and No RTW (N = 22). RTW trajectories differed consistently with regard to previous sick-leave duration and social support at work. More unique predictors of RTW trajectories included gender, rewards at work, work performance impairment due to health problems, home-to-work interference and stress-related exhaustion disorder. CONCLUSIONS The study may have important clinical implications for identifying patients belonging to a particular RTW trajectory. Knowledge on the modifiable work environment factors that differentiated between the RTW trajectories could be useful for designing effective workplace interventions, tailored to particular needs of employees with CMDs. However, in a first step, the results need to be replicated.