Objective To characterize changes in sleep disturbance in women with breast cancer during multiple cycles of chemotherapy, to explore potential classes of women experiencing similar patterns of disturbed sleep and to identify demographic and disease-related predictors of these patterns.Methods A total of 219 women with breast cancer receiving chemotherapy were surveyed by using a self-designed general information questionnaire and the Patient-reported Outcomes Measurement System-Breast-Chemothe-rapy sleep disorders instrument before the first cycle of chemotherapy and on the 7th day after completion of the first cycle, 2-4 cycles, and the last cycle of therapies. Repeated measures ANOVA was used to test changes of sleep disorder over time. Latent class growth analysis was performed to examine sleep disturbance trajectories. Chi square test and logistic regression were used to explore the predictors of trajectories. Results The scores of sleep disorders were 47.50±8.82,51.38±10.83,50.66±9.90, and 50.46±9.99 at different time points, showing significant difference (P<0.01). Four sleep disturbance classes were identified: normal group (19.2%), mildly disturbed group (25.1%), moderately disturbed group (50.7%), and highly disturbed group (5.0%). The sleep disturbance trajectories were predicted by regional lymph node metastasis （OR=1.475，P=0.042）, receiving targeted therapy （OR=2.135，P=0.011）,and participation in clinical trials （OR=1.899，P=0.040）．Conclusion The sleep disorder level changes over the course of treatment in breast cancer patients receiving chemotherapy, and the longitudinal trajectories have significant group heterogeneity. Regional lymph node metastasis, use of targeted therapy and participation in clinical trails can predict the trajectories of sleep disorders. Medical staff can use these findings to identify patients at moderate to high risk of sleep disturbance, and provide whole-process management and precision care.