The　increasing　investment　in　environmental　pollution　management　urgently　needs　the　scientific　utilization　of　environmental　management　costs.　However,　three　challenges　must　be　addressed　in　environmental　pollution　management.　First,　the　reliability　of　environmental　pollution　data　is　often　disregarded,　which　may　produce　unreliable　inference　results.　Second,　there　are　many　uncertainties　in　actual　practice,　which　are　neglected　in　existing　cost　prediction　and　efficiency　evaluation　of　environmental　pollution　management.　Third,　existing　research　studies　mainly　focused　on　either　efficiency　evaluation　or　cost　prediction　and　ignored　the　importance　of　combining　both　for　environmental　pollution　management.　To　address　these,　an　extended　belief　rule　base　(EBRB)　model　that　considers　consequence　reliability　is　proposed　to　predict　the　interval　costs,　followed　by　an　interval　data　envelopment　analysis　(IDEA)　model　that　considers　undesirable　output　to　evaluate　interval　efficiencies　of　environmental　pollution　management.　Based　on　these　improved　models,　an　integrated　model　named　as　EBRB-IDEA　model　is　further　developed　under　interval　uncertainty.　To　verify　its　practical　usage,　the　environmental　pollution　data　of　29　Chinese　provinces　from　2004　to　2017　were　used　to　carry　out　a　case　study.　The　experimental　results　demonstrated　that　the　EBRB-IDEA　model　did　not　only　achieve　the　desired　interval　prediction　costs　and　efficiency　evaluation　but　also　effectively　distinguished　regional　differences　in　the　efficiency　of　environmental　pollution　management　compared　with　existing　models.