For　the　influence　factors＇　complexity　of　urban　post-earthquake　fire,　the　attribute　recognition　theory　and　the　evidence　theory　were　proposed　for　the　potential　risk　analysis　of　urban　post-earthquake　fire.　First,　the　model　of　attribute　recognition　based　on　the　principle　of　maximum　entropy　was　established　to　eliminate　all　kinds　of　uncertainty　to　max;　secondly,　the　evaluation　indicators　for　urban　post-earthquake　fire　were　presented　in　terms　of　urban　disaster　hazard,　exposure　and　vulnerability　of　disaster　bearing　body,　disaster　prevention　and　mitigation　capability.　Meanwhile,　the　grading　standards　of　evaluation　index　were　given;　finally,　potential　risk　analysis　for　post-earthquake　fire　within　a　city　was　provided　to　validate　its　efficiency　and　feasibility　by　the　evidence　theory.　The　results　indicate　that　the　middle　and　higher　risk　area　will　reach　50.20%,　61.00%　and　64.90%　respectively　when　VI,　VII,　VIII-degree　earthquake　occurs.　In　this　study,　we　verify　the　application　of　attribute　recognition　and　evidence　theory　for　potential　risk　assessment　of　urban　post-earthquake　fire　and　provide　a　measurable　and　achievable　approach　for　the　potential　risk　assessment　of　urban　post-earthquake　fire.