Environmental　governance　cost　prediction　is　an　essential　process　in　environmental　protection.　However,　the　existing　environmental　governance　cost　prediction　methods　are　facing　two　challenges:　First,　the　principal　components　of　environmental　indicator　information　must　be　accurately　extracted　without　considering　the　independence　of　environmental　indicators.　Second,　the　higher　interpretability　and　the　lower　complexity　must　be　taken　into　account　with　the　desired　accuracy　for　improving　the　cost　prediction　of　environmental　governance.　Therefore,　the　fuzzy　rule　based　system　(FRBS)　and　feature　extraction　are　introduced　to　propose　a　new　environmental　governance　cost　prediction　method,　named　FRBS-FE,　in　which　the　feature　extraction　is　used　to　extract　the　principal　components　of　environmental　indicator　information　firstly,　and　then　all　these　principal　components　are　applied　to　generate　a　FRBS-FE　for　better　environmental　governance　cost　prediction.　A　case　study　involving　29　provinces　of　China　is　carried　out　to　demonstrate　the　effectiveness　of　the　FRBS-FE.　The　results　showed　that　the　FRBS-FE　not　only　can　accurately　predict　different　kinds　of　environmental　governance　costs,　but　also　have　superior　performance　in　comparison　with　previous　cost　prediction　methods.