AIM: To describe tyrosine kinase inhibitor (TKI) treatment patterns and analyze co-variates of TKI switch for chronic myeloid leukemia (CML) patients in a center from China. METHODS: A retrospectively study was designed to analyze TKI switching patterns, reasons and associated covariates in patients with CP-CML. RESULTS: 1766 patients receiving initial imatinib (n = 1374), nilotinib (n = 254), dasatinib (n = 63) and flumatinib (n = 75) therapy were retrospectively interrogated. Median follow-up was 48 (IQR, 24-77) months. TKI switch proportions were 32% (570/1766) for first-line, 36% (208/570) for second-line and 34% (71/208) for third-line. Common therapy sequences included imatinib-dasatinib (37%) or nilotinib (35%) in those with 1 switch, imatinib-nilotinib-dasatinib (25%) with 2 switches and imatinib-nilotinib-dasatinib-olverembatinib (18%) with 3 switches. TKI switches were mainly due to resistance (64%, 76%, 88% across lines) and intolerance (19%, 14%, 7%). Multivariable analyses revealed ELTS intermediate/high-risk group (vs. low-risk), male, and lower hemoglobin were significantly associated with a higher probability of TKI switch. Compared to imatinib, initial nilotinib or dasatinib had lower switch rates. Male and ELTS high-risk (vs. low/intermediate) were associated with resistance-related switches, while lower hemoglobin, older age and initial dasatinib or flumatinib (vs. imatinib) were associated with intolerance-related switches to second-line therapy. Second-line imatinib/flumatinib (vs. nilotinib/dasatinib) and no/nonspecific ABL mutation were associated with resistance-related switches to third-line therapy. CONCLUSION: These findings emphasized the complexities involved in the management of patients with CP-CML and highlighted the importance of personalized treatment strategies.