This single-center observational study aimed to assess the inter-rater reliability (IRR) of Kampo medicine pattern diagnosis, which is modularized into three modules for chronic diseases, using 64 participants' information documents. The linearly weighted percentage of agreement and Gwet's agreement coefficient (AC) 2 for the deficiency-excess module, among three specialists, were 85.9% and 0.708, respectively. The unweighted percentage of agreement and Gwet's AC1 were 64.6% and 0.542 for the cold-heat, and 35.9% and 0.254 for the qi-blood-fluid modules, respectively. Our findings suggest that our modularization method may improve the IRR of Kampo medicine pattern diagnosis.