Abstract Background The China-Gates TB project Phase II implemented case-based payment reform in three Chinese counties in 2014, designed specifically for patients diagnosed with Tuberculosis (TB). This study aimed to examine the changes in utilization and expenses of outpatient services before and after the reform implementation, among TB patients in the three counties in China. Methods We collected quantitative data using surveys in 2013 (baseline year) and 2015 (final year). We used outpatient hospital records to measure service utilization and medical expenses of TB patients. We conducted qualitative interviews with local health authorities, officers of health insurance agencies, and hospital managers (n = 18). We utilized three focus group discussions with hospital staff and TB doctors and nurses. The χ² tests and Mann-Whitney U tests were used to analyse quantitative data, and the thematic analysis using a framework approach was applied to analyse qualitative data. Results Dantu and Yangzhong counties enacted TB-specific case-based payment method in 2014. Jurong County maintained global budget payment but raised the reimbursement rate for TB care. Compared to the baseline, the percentage of TB patients in Dantu and Yangzhong with eight or above outpatient visits increased from 7.5 to 55.1% and from 22.1 to 53.1% in the final survey, respectively. Jurong experienced the opposite trend, decreasing from 63.0 to 9.8%. In the final survey, the total outpatient expenses per patient during a full treatment course in Dantu (RMB 2939.7) and Yangzhong (RMB 2520.6) were significantly higher than those in the baseline (RMB 690.4 and RMB 1001.5, respectively), while the total outpatient expenses in Jurong decreased significantly (RMB 1976.0 in the baseline and RMB 660.8 in the final survey). Health insurance agencies in Dantu and Yangzhong did not approve the original design with outpatient and inpatient expenses packaged together, revealed by qualitative interviews. Furthermore, staff at designated hospitals misunderstood that health insurance agencies would only reimburse actual expenses. Many TB doctors complained about their reduced salary, which might be due to decreased hospital revenue generated from TB care after the payment method reform. Conclusions The intended effect on cost containment of case-based payment was not evident in Dantu and Yangzhong. In Jurong, where the global budget payment system maintained with the reimbursement rate enhanced, we found an effect on cost containment, but the quality of TB care might be compromised. The TB-specific case-based payment method could be redesigned to combine payment on outpatient and inpatient expenses and to set an appropriate payment standard for TB care during a full treatment course. Local health insurance agencies have to provide explicit explanations on the payment method. TB care providers should be provided with proper incentives. Monitoring and evluaiton on the quality of TB care should be undertaken at regular intervals.