Traditional Chinese medicine（TCM） medical institutions are the leading providers of TCM services in China，and their service efficiency directly affects the overall quality of TCM services nationwide. This study adopts the undesirable Slacks-Based Measurement（SBM）model to measure the service efficiency of TCM medical institutions across 30 provinces in China from 2012 to 2021. On this basis，it explores the regional differences，dynamic evolution，and convergence characteristics using the Dagum Gini coefficient， kernel density estimation，and β-convergence model. The results show that the overall service efficiency of TCM medical institutions in China remained at approximately 0.96 from 2012 to 2021，with average efficiencies of 1.10， 0.90，0.94，and 0.66 in the eastern，central，western and northeastern regions，respectively. The Dagum Gini coefficient decreased from 0.221 to 0.140 during this period，indicating that the primary source of regional difference shifted from inter - regional differences to hyper - variable density. Within the eastern，central，and western regions， the service efficiency differences of TCM medical institutions narrowed with the multipolar trend gradually weakening，while the service efficiency differences within the Northeast region have been enhanced. There are Absolute and conditional β convergence observed across all regions of China，along with a significant regional heterogeneity.