Against　the　background　of　deepening　financial　reform,　the　improvement　of　operational　efficiency　has　be　-come　the　consensus　of　the　banking　industry.　However,　the　impact　of　undesirable　outputs　should　also　be　considered　in　the　efficiency　evaluation　of　commercial　banks,　for　which　the　directional　distance　function　(DDF)　is　a　feasible　method.　The　present　study　proposes　the　application　of　a　multi-period　leader-follower　model　to　dynamically　evaluate　the　operational　efficiency　of　commercial　banks.　The　proposed　model　over-comes　the　defect　of　ignoring　the　internal　structure　of　the　decision　making-unit　(DMU)　present　in　the　traditional　data　envelopment　analysis　(DEA)　method,　and　expands　the　leader-follower　model　with　single　-period　efficiency.　In　this　study,　the　following　four　observations　are　made　from　the　analysis　of　the　effi-ciency　results　of　16　representative　commercial　banks　from　2009　to　2018.　(1)　The　multi-period　leader-follower　model　is　more　distinguishable　in　terms　of　efficiency　calculation.　(2)　During　the　period　of　2014-　2018,　the　stage　and　system　efficiencies　were　improved　and　more　stable　than　those　during　the　period　of　2009-2013.　(3)　From　2009　to　2018,　the　inefficiency　value　of　the　average　fundraising　efficiency　was　higher　than　that　of　the　average　capital-utilization　efficiency.　(4)　The　system　and　capital-utilization　efficiencies　of　state-owned　commercial　banks　are　inferior　to　those　of　joint-stock　banks.　Furthermore,　the　fundraising　efficiency　of　state-owned　commercial　banks　is　not　much　less　than　that　of　joint-stock　banks.　The　domes　-tic　and　international　macro　and　micro　influencing　factors　of　commercial　bank　efficiency　are　analyzed.　Finally,　some　policy　recommendations　for　improving　the　operational　efficiency　of　commercial　banks　are　put　forward.(c)　2022　Elsevier　Ltd.　All　rights　reserved.