Efficient　allocation　of　carbon　dioxide　(CO2)　emissions　allowances　is　essential　to　the　realization　of　the　emissions　reduction　goal.　Using　China　as　an　example,　this　study　applies　a　zero　sum　gains–data　envelopment　analysis　(ZSG-DEA)　model　to　translate　the　country＇s　reduction　target　into　provincial　goals　based　on　the　Chinese　government＇s　2020　reduction　goal.　To　evaluate　efficiency　scores,　we　employ　the　Epsilon-based　measure　model　that　combines　radial　and　non-radial　characteristics.　Different　from　prior　studies,　we　emphasize　the　effect　of　industrial　structure　on　CO2　emissions　and　calculate　a　composite　gross　domestic　product　(GDP)　index　as　one　of　the　output　variables.　By　utilizing　the　ZSG-DEA　model　and　based　on　China＇s　provincial　CO2　emissions　allowances　as　the　input　variable,　the　composite　GDP　index,　and　the　population　as　the　output　variable,　we　obtain　an　optimal　reallocation　scheme　after　four　iterations.　A　comparison　between　the　results　and　official　reduction　goals　shows　that　the　government　should　increase　CO2　emissions　allowances　appropriately　for　most　provinces　with　high　proportions　(more　than　40%)　of　second　industry　increment　in　GDP,　thereby　carrying　implications　for　policymakers.　With　China　as　an　example,　these　findings　have　global　implications　as　regards　developing　a　more　scientific　allocation　scheme,　especially　for　regions　with　unbalanced　industrial　structures.　©　2019　Elsevier　Ltd