The　projection　pursuit　evaluation　model　is　applied　to　study　the　problem　of　multi-factor　evaluation　of　urban　comprehensive　disaster-carrying　capability.　Projection　pursuit　can　project　high　dimensional　data　to　low　dimensional　space,　and　meanwhile　particle　swarm　optimization　algorithm　is　used　to　optimize　its　projection　direction.　Through　studying　the　main　characteristics　in　1-D　space,　the　key　information　in　the　original　data　can　be　obtained,　which　can　be　used　to　classify　or　evaluate　urban　comprehensive　disaster-carrying　capability.　In　addition,　with　application　of　this　method,　the　subjectivity　of　safety　factors　can　also　be　effectively　eliminated.　The　results　of　case　study　show　that　the　analysis　results　are　completely　consistent　with　the　actual　situation,　so　the　method　can　construct　a　characteristic　index,　and　the　main　impact　effects　can　be　analyzed　according　to　the　result.　The　model　is　qualified　with　simplicity　and　can　provide　a　new　approach　for　evaluating　urban　disaster-carrying　capability.