PURPOSE: Depression rating scales have predicted long-term sickness absence (LTSA) in previous studies. With this study, we investigated to what extent single symptoms from a depression rating scale predicted LTSA among employees who were free of clinical depression. METHODS: We studied 6,670 female employees in the Danish eldercare sector. Frequency of 12 depressive symptoms over the last 2 weeks was assessed with the Major Depression Inventory. A symptom was considered as elevated if it was present at least "slightly more than half of the time." Data were linked to a national register on LTSA (≥3 weeks). We calculated hazard ratios (HR) from Cox's proportional hazard models to analyze whether a symptom predicted time to onset of LTSA during a 1-year follow-up. Analyses were adjusted for age, family status, health behaviors, occupational group, and previous LTSA. RESULTS: Of the 12 symptoms, three predicted LTSA after adjustment for covariates: "felt low in spirits and sad" (HR = 1.41, 95 % CI = 1.05-1.89), "felt lacking in energy and strength" (HR = 1.33, 95 % CI = 1.08-1.64), and "had trouble sleeping at night" (HR = 1.38, 95 % CI = 1.09-1.74). CONCLUSION: Among female eldercare workers free of clinical depression, feelings of low spirits and sadness, feelings of lack of energy and strength, and sleep disturbances predict risk of LTSA. Interventions that decrease the prevalence of these symptoms might contribute to a reduction in LTSA in this population.