OBJECTIVE: Breast cancer incidence among Chinese females has been increased for decades. In order to evaluate the association between sleep duration and female breast cancer incidence, Kailuan prospective cohort was applied as the basis of this study. METHODS: Among females in Kailuan health detection program during 2006-2011, baseline information such as demography, sleeping habits, height and weight were collected, as well as the information of newly-diagnosed breast cancer cases during the follow-up period. Cox proportional risk regression model was applied to evaluate the risk ratio of sleep duration and female breast cancer incidence.RESULTS: A total of 24 692 women were recruited in this study, and there were 108 029.22 person-years of follow up and 4.38 years of average follow-up duration. During the follow-up period, 107 breast cancer were newly indentified in this cohort. Compared with females with 7-hour sleep duration, females with insufficient sleep (≤6 h) or longer sleep (≥8 h) duration showed no significant relation to breast cancer incidence (HR=0.99, 95%CI=0.52-1.87; HR=0.84, 95%CI=0.49-1.45), after adjusting with age, education level, alcohol drinking, and the other confounding factors, the results also showed no statistical association (insufficient sleep: HR=0.82, 95%CI=0.43-1.56; longer sleep: HR=0.94, 95%CI=0.54-1.64). Similar results were obtained in subgroup analysis, which stratified by age, snoring, or BMI, and sensitivity analysis. CONCLUSION: Our study don't support that sleep duration is significantly associated with the risk of female breast cancer.