Environmental　governance　cost　prediction　can　avoid　blind　investment　and　waste　of　resources　and　achieve　effective　cost　planning　for　sustainable　development　of　resources　and　environment.　For　the　sake　of　solving　the　problem　that　most　previous　studies　failed　to　consider　the　causal　relationship　and　data　reliability　of　environmental　governance　inputs　and　outputs,　a　new　environmental　governance　cost　prediction　method　is　proposed　under　the　framework　of　the　evidential　reasoning　(ER)　rule　with　three　improvements　comparing　to　existing　methods:　(1)　the　causal　relationship　of　environmental　governance　inputs　and　outputs　is　embedded　into　evidence　representation　for　better　extracting　knowledge　from　data;　(2)　the　efficiency　about　the　minimum　inputs　to　achieve　the　maximum　outputs　is　used　to　evaluate　the　data　reliability　of　environmental　governance　inputs　and　outputs;　and　(3)　a　new　analytical　ER　rule　is　investigated　to　optimize　the　process　of　evidence　combination.　Hence,　the　new　method　includes　the　calculation　of　belief　distributions,　evidence　reliabilities,　and　evidence　weights,　as　well　as　the　combination　of　evidences　to　predict　environmental　governance　costs.　In　the　case　study,　the　data　of　30　provinces　in　Mainland　China　from　2005　to　2020　are　collected　to　verify　the　effectiveness　of　the　new　method.　Results　show　a　high　level　of　accuracy　of　the　new　method　over　other　existing　methods.　©　2023,　The　Author(s),　under　exclusive　licence　to　Springer-Verlag　GmbH　Germany,　part　of　Springer　Nature.