The　traditional　three-stage　data　envelopment　analysis　(DEA)　model　only　measures　exact　input–output　indicator　data,　but　cannot　perform　efficiency　analysis　on　uncertain　data.　The　interval　DEA　method　does　not　exclude　the　influence　of　external　environmental　factors.　Therefore,　this　paper　combines　the　traditional　three-stage　DEA　model　with　the　interval　DEA　method,　and　proposes　a　three-stage　interval　DEA　efficiency　model,　which　eliminates　the　impact　of　external　environmental　factors　and　real-izes　the　measurement　of　the　efficiency　for　interval　data.　From　the　perspective　of　the　impact　of　environmental　factors,　defining　the　degree　of　efficiency　change　vector,　a　clustering　analysis　technique　based　on　the　efficiency　change　degree　vector　is　proposed　to　provide　improvement　benchmark　for　poorly　performing　decision-making　units.　Finally,　an　example　is　used　to　demonstrate　the　feasibility　and　validity　of　the　proposed　method　in　this　paper.　©　2021　The　Authors.　Published　by　Atlantis　Press　B.V.