This study aimed to clarify the difference between the durations of first and second periods of depression-induced sick leave and to identify predictors of a prolonged second period of depression-induced sick leave. Among Japanese employees who were registered in the Japan sickness absence and return to work (J-SAR) study, the subjects were those employees who returned to work after an initial period of depression-induced sick leave (F3; ICD-10, based on a psychiatrist’s certificate), and returned to work after a second period of depression-induced sick leave. The subjects’ second periods of sick leave (mean: 156.9 d) were longer than their first periods of sick leave (107.3 d) (Wilcoxon test, p =0.007). In the logistic regression analysis (Table 2), “longer duration of the first period of sick leave” (Odds ratio: 3.258, 95%CI: 1.780–5.963, p <0.001) was identified as a significant predictor of a longer recurrent period of sick leave. Individuals who experience a long initial period of depression-induced sick leave should be supported carefully by occupational health professionals after they RTW.