Abstract: OBJECTIVE：This study aimed to develop a method for assessing uncertainty in microbial testing of medical devices using the black-box approach，and to bolster integrity reliability of test outcomes. METHODS：The black-box method and the reproducibility standard deviation formula were used to determine combined uncertainty in microbial testing for medical devices. Parallel experiments were performed on domestically produced transcutaneous electrical nerve stimulators and recombinant collagen medical products to evaluate the effects of varying operators，sampling techniques，dilution factors，and reading methods on the total bacterial count for different samples. RESULTS：The black-box method was both convenient and practical in facilitating the calculation of combined uncertainty. It enabled the dynamic updating of laboratory data in response to new uncertainties and data. For the transcutaneous nerve stimulator，the synthetic standard uncertainty was 3.69，with an expanded uncertainty of 7.38，reporting a final detection result of (54.8±7.38) CFU/mL. For the recombinant collagen medical product，the synthetic standard uncertainty was 2.42，with an expanded uncertainty of 4.84，reporting a final detection result of (20.8±4.84) CFU/mL. The synthetic uncertainty for in-laboratory bacterial detection was 8.81. CONCLUSION：The black-box method effectively evaluated and reported the uncertainty in microbial testing for medical devices，which is crucial for improving test quality，compliance assessment，and the mutual recognition of results across laboratories.

Key words: medical devices, microbial testing, uncertainty assessment, black-box method, reproducibility standard deviation