Transformer　oil-paper　insulation　aging　is　an　important　factor　affecting　the　normal　operation　of　transformers.　At　present,　the　depolarization　current　(PDC)　method　based　on　the　time　domain　detection　is　widely　used　in　transformer　insulation　condition　diagnosis,　but　the　feature　extraction　of　depolarizing　current　line　is　still　in　the　initial　stage.　According　to　the　dielectric　response　and　trap　theory,　we　put　forward　a　method　to　draw　the　depolarizing　trap　density　spectrum,　extract　two　features　including　peak　value　Smax　and　peak　time　constant　Tmax,　and　then　use　them　to　evaluate　the　state　of　transformer　insulation.　Firstly,　different　polarization　voltage　tests　are　used　to　analyze　the　relationship　between　new　characteristic　value　and　polarization　voltage.　Secondly,　the　trap　density　of　transformers　with　different　furfural　concentration　are　comparatively　analyzed.　Finally,　a　large　number　of　instances　on　oil-paper　insulation　aging　diagnosis　are　verified　and　analyzed.　The　research　results　show　that　the　peak　value　is　positively　related　with　furfural　content,　and　peak　time　constant　is　negatively　related　with　furfural　content.　With　the　deepening　of　oil-paper　insulation　aging,　the　peak　value　Smax　increases　gradually,　but　the　peak　time　constant　Tmax　decreases　gradually.　The　value　range　of　two　characteristics　proposed　can　be　used　as　the　basis　of　the　diagnosis　of　transformer　insulation　aging　status.　©　2017,　High　Voltage　Engineering　Editorial　Department　of　CEPRI.　All　right　reserved.