It　is　essential　to　establish　a　dielectric　response　mathematical　model　and　identify　its　parameters　exactly　for　assessing　the　aging　state　of　oil-paper　insulation.　Consequently,　we　adopted　the　line　shape　factor　of　microscopic　dynamics　to　establish　a　non-typical　line　shape　dielectric　response　function　for　the　depolarization　process　not　being　reflected　by　conventional　model　accurately.　The　second　time-differential　analysis　was　firstly　proposed　to　be　applied　in　decomposition　of　different　relaxation　processes.　This　method　not　only　determined　branches＇　polarization　number　accurately,　but　also　ensured　the　parameter　identification　uniquely　by　spectral　line　characteristics　extracted.　Finally,　the　proposed　method　was　used　to　identify　the　parameters　of　depolarizing　current　curve　measured　from　a　220　kV　transformer.　The　results　show　that　the　depolarizing　curve　by　the　improved　dielectric　response　model　and　the　second　time-differential　analysis　has　better　coincidence　of　97.58%　with　the　test　curve　while　the　conventional　one　is　88.56%,　which　verifies　the　fitness　and　accuracy　of　the　method　on　the　parameter　identification　of　oil-paper　insulation　dielectric　response.　©　2017,　High　Voltage　Engineering　Editorial　Department　of　CEPRI.　All　right　reserved.