This　paper　is　based　on　the　generalized　regression　neural　network(GRNN)　theory,　finding　the　rules　of　data　by　it,　constructing　prediction　model　of　GRNN,　then　use　it　to　forecast　the　life　of　oil　paper　insulation　transformers.　In　the　process　of　aging,　insulating　paper　of　a　transformer　will　generate　feature　products　such　as　furfural,　CO2　and　CO.　The　content　of　each　product　and　the　corresponding　time　parameter　is　used　as　the　model　input.　The　test　samples　are　collected　as　the　basic　data,　to　predict　the　life　of　corresponding　transformers　by　using　this　model.　Comparison　of　the　results　of　GRNN　and　the　measured　data　show　that　the　results　of　GRNN　are　in　accordance　with　the　measured　values,　which　proves　the　rationality　of　the　model.　And　it　is　also　significant　for　the　further　study　of　material　aging　status　of　insulation　monitoring.　©,　2014,　Xi＇an　High　Voltage　Apparatus　Research　Institute.　All　right　reserved.