Objective The aim of this study was to determine whether the frequency of changing work shift and overtime are associated with registered nurses’ menstrual characteristics.

Method Female nursing staff aged 20–45 years and working at hospitals were invited to participate in the study. Menstrual characteristics and work shifts were recorded daily for 180 days. A multivariate logistic model was performed. Important confounders such as age, educational level, occupational category, regular excise habit, and major earner of family were adjusted.

Result A total of 330 nurses and 1,437 menstrual cycles were collected and eligible for the final analysis. The adjusted odds ratio for shorter bleeding time (≦3 days) was associated with changing work shift (involved one night shift) more than 3 times between menstrual cycles were 2.2 (95% CI = 1.1–4.3). The adjusted odds ratios for longer menstrual cycle lengths (>40days) due to changing work shift (involved one night shift) more than 3 times between menstrual cycles were 4.7(95% CI = 3.1–7.1). The adjusted odds ratios for dysmenorrhea due to overtime more than 40 hours during 28 days before menstrual were 2.9 (95% CI = 1.6–5.2).

Conclusion The high frequency of changing work shift (esp. from day shift change to night shift or from evening shift to night shift) and overtime (>40 hours/ 28days) may affect childbearing aged female nurses’ reproductive function.