Objective To define an objective evaluation model for metadata integrity of randomized controlled trials (RCTs) in traditional Chinese medicine (TCM), and to evaluate the data integrity of RCT reports published in TCM journals. Methods Retrieving Chinese medicine RCT literature and extracting data, using the metadata specification list and customized evaluation model defined in the project "Intelligent Construction and Application Demonstration of the Evidence System of Chinese Medicine Dominant Diseases" to analyze RCTs from the perspective of data integrity. Results A metadata interface specification and an objective evaluation model for RCT metadata integrity were proposed. A total of 37 361 articles of 10 diseases from 1986 to 2020 were evaluated. Among them, 6 743 reports failed to meet the basic requirements of metadata specifications. The proportion of reports with no missing required items was between 73% and 97%. "tcm_disease" and "num_drop_total" had a greater impact on completeness for the required items. The reporting rates of the items in the "age_sd" and "history_sd" in the "group" section, and "dosage", "dosage_form" and "dosage_freq" in the "interventions" section were low. The average score of RCT report was 71.39 points. Conclusions There is room for improvement in the integrity of RCT data in TCM, and data reporting is urgently required to be standardized. The metadata specification and completeness objective evaluation model proposed in this study can provide references for improving the data integrity of clinical trial reports of TCM.