Depressive symptoms are a serious problem in workplaces. Hospital staff members, such as newly licensed registered nurses (NLRNs), are at particularly increased risk of these symptoms owing to their limited experience. Previous studies have shown that a brief program-based cognitive behavioral therapy program (CBP) can offer effective treatment. Here, we conducted a longitudinal observational study of 683 NLRNs (CBP group, n = 522; no-CBP group, n = 181) over a period of 1 year (six times surveys were done during this period). Outcomes were assessed on the basis of surveys that covered the Beck Depression Inventory-I (BDI). The independent variables were CBP attendance (CBP was conducted 3 months after starting work), personality traits, personal stressful life events, workplace adversity, and pre-CBP change in BDI in the 3 months before CBP (ΔBDIpre-CBP). All factors were included in Cox proportional hazards models with time-dependent covariates for depressive symptoms (BDI ≥10), and we reported hazard ratios (HRs). Based on this analysis, we detected that CBP was significantly associated with benefit for depressive symptoms in all NLRNs (Puncorrected = 0.0137, HR = 0.902). To identify who benefitted most from CBP, we conducted a subgroup analysis based on the change in BDI before CBP (ΔBDIpre-CBP). The strongest association was when BDI scores were low after starting work and increased before CBP (Puncorrected = 0.00627, HR = 0.616). These results are consistent with previous findings, and indicate that CBP may benefit the mental health of NLRNs. Furthermore, selective prevention based on the pattern of BDI change over time may be important in identifying who should be offered CBP first. Although CBP is generally effective for all nurses, such a selective approach may be most appropriate where cost-effectiveness is a prominent concern.