The　traditional　eigenvalue　buckling　mode　method　used　for　the　stability　analysis　of　suspended　domes　has　been　deemed　unsafe　and　unreasonable.　A　random　imperfection　method　is　proposed　in　this　paper　that　is　suitable　for　the　stability　analysis　of　suspended　domes.　Using　this　method,　the　impact　of　initial　imperfections,　such　as　joint　position　deviation,　eccentricity　of　members,　variation　in　cable　pretension　and　elastic　modulus　of　materials,　on　the　stability　of　nonlinear　geometries　as　well　as　the　stability　of　both　nonlinear　geometries　and　materials　(i.e.,　ultimate　bearing　capacity)　are　studied.　The　analysis　results　using　this　method　are　compared　to　those　obtained　from　the　traditional　eigenvalue　buckling　mode　method.　This　study　indicates　that　the　random　imperfection　method　can　more　reasonably　demonstrate　the　impact　of　initial　imperfections　on　the　stability　of　suspended　domes.　Additionally,　the　shape　and　size　of　the　initial　imperfections　applied　by　the　random　imperfection　method　are　more　reasonable,　and　the　results　are　safer.　The　random　imperfection　method　is　successfully　used　in　the　stability　analysis　of　a　large　span　suspended　dome,　and　the　results　of　this　method　are　nearly　equivalent　to　that　of　the　model　test,　which　verifies　the　method.