Urban　collaboration　serves　as　a　potentially　efficient　policy　tool　in　advancing　metropolitan　governance.　However,　few　empirical　evidences　have　been　devoted　to　illustrate　why　some　urban　collaborations　can　persist　over　a　longer　term.　In　this　article,　we　draw　upon　the　institutional　collective　action　framework　and　hypothesize　that　actor　heterogeneity,　political　institutions,　and　community　characteristic　will　jointly　affect　the　sustainability　performance　of　urban　collaboration.　The　combined　effect　is　examined　by　using　fuzzy-set　qualitative　comparative　analysis　with　45　intra-provincial　Urban　Agglomeration　Arrangements　from　China.　The　results　indicate　that　vertical　government　intervention　plays　as　a　vital　political　factor　within　the　Chinese　context,　and　four　configuration　paths　can　illustrate　the　persistence　of　collaboration:　exchange　effect　by　homogeneous　actors,　substitutive　effect　by　geographical　proximity　and　vertical　government　intervention,　complement　effect　by　vertical　government　intervention,　and　covariant　effect　by　city　hierarchy　heterogeneity　and　group　size.　The　findings　contribute　to　bridging　the　knowledge　gap　between　urban　collaboration　and　its　sustainability　performance.　©　2025　Elsevier　Ltd