Evaluating　and　improving　the　performance　of　green　development　(GD)　is　essential　for　the　government　to　formulate　sustainable　development　policies.　This　study　first　analyzes　the　connotation　and　structures　of　GD　from　systemic　perspective.　Secondly,　based　on　the　two-stage　eco-inefficiency　(TE)　model,　non-cooperative　theory,　the　correlation　coefficient　and　standard　deviation,　and　fuzzy　set　qualitative　comparative　analysis　method　(fsQCA),　an　improved　integrated　method,　i.e.,　ITNE-fsQCA,　is　proposed.　Thirdly,　the　validity　and　accuracy　of　ITNE-fsQCA　are　verified　using　the　data　from　30　Chinese　provinces　in　2020,　and　the　results　are　compared　with　those　obtained　from　TE　and　two-stage　non-cooperative　eco-inefficiency　(TNE).　Finally,　the　driving　types　and　configuration　pathway　are　discussed　based　on　ITNE-fsQCA.　The　results　show　that:　(1)　compared　to　TE　and　TNE,　ITNE-fsQCA　helps　to　avoid　the　overestimation/underestimation　of　inefficiency　in　complex　systems;　(2)　there　are　8　configuration　pathways　that　can　effectively　improve　GD　performance　in　30　Chinese　provinces　in　2020,　and　these　pathways　are　categorized　into　4　driving　types;　(3)　the　enhancement　of　green　technology　innovation　plays　a　crucial　role　in　provincial　sustainable　development;　and　(4)　the　provincial　GD　inefficiencies　in　China　are　manifest　various　combinatorial　forms.　These　findings　provide　the　government　with　innovative　ideas　and　methodologies　for　formulating　sustainable　development　policies.　©　2024　Elsevier　Ltd