Innovative　industrial　clusters　are　regarded　as　regional　innovation　centers　and　have　become　an　important　support　for　the　national　innovation　strategy　in　China.　However,　in　the　actual　development　process　of　innovative　industrial　clusters,　there　are　problems　of　unbalanced　development　and　low　development　efficiency.　This　study　selected　the　data　envelopment　analysis　(DEA)-Malmquist　model　to　analyze　the　development　efficiency,　spatiotemporal　evolution　characteristics,　and　spatial　improvement　of　China＇s　innovation　industrial　clusters.　This　study　is　different　from　previous　studies　that　used　a　single　industrial　cluster　as　a　decision-making　unit.　Instead,　it　uses　provinces　as　a　decision-making　unit.　In　calculation　and　analysis,　provinces　are　used　as　decision-making　units　instead　of　single　industrial　clusters　as　decision-making　units　in　previous　studies.　The　results　showed　the　following:　(1)　The　average　development　efficiency　of　the　innovative　industrial　clusters　was　0.652,　which　was　relatively　low　and　had　great　potential　for　effective　development.　Meanwhile,　the　average　value　of　the　Malmquist　index　was　0.932,　which　declined　at　an　average　annual　rate　of　7.5%　during　the　study　period,　mainly　due　to　the　slowdown　in　technological　progress　and　to　the　decline　in　pure　technical　efficiency.　(2)　There　were　no　significant　differences　in　the　development　efficiency,　Malmquist　index,　or　decomposition　indices　of　China＇s　innovation　industry　among　the　four　regions　investigated,　indicating　that　regional　advantages　have　not　been　significantly　exerted.　(3)　Through　further　analysis　of　the　potential　for　improvement,　a　quantitative　improvement　of　the　space　of　the　input-output　factors　of　China＇s　19　DEA-inefficient　provinces　was　obtained.