T-peel　test　of　ductile　adhesive　is　always　an　interesting　and　important　topic　in　peeling　mechanics　and　fracture　mechanics.　In　this　paper,　the　pre-notch　and　pre-crack　effect　on　the　T-peel　test　with　a　ductile　adhesive　layer　is　studied.　Based　on　the　experimental　test　of　SAC　305　adhesive　and　copper　adherent　T-peel　tests,　it　is　found　that　the　peak　loads,　fracture　toughness,　resistance　curve,　and　adhesive　toughness　are　sensitive　to　pre-crack　or　pre-notch　lengths.　Short　pre-notch　or　pre-crack　trigger　unsteady　state　debonding　frequently,　and　deep　pre-notch　or　pre-　crack　conditions　achieve　steady　state　debonding　more　easily.　Different　data　reduction　schemes　such　as　simple　beam　theory　(SBT),　corrected　beam　theory　(CBT),　and　compliance-based　beam　method　(CBBM)　will　lead　to　variations　of　fracture　toughness　extractions　for　T-peel　tests.　Trapezoidal　shape　cohesive　zone　model　(CZM)　is　found　to　be　easier　matching　T-peel　tests　conducted　in　this　paper.　However,　T-peel　samples　with　shorter　pre-notch　or　pre-crack　lengths　are　hard　to　simulate　with　the　CZM　method　due　to　the　unstable　crack　propagation.　The　reasons,　which　may　lead　to　the　discrepancy　between　pre-notch　and　pre-crack　T-peel　samples,　are　also　presented　and　discussed.　The　selection　of　different　pre-notch　or　pre-crack　lengths　is　a　very　important　factor　that　needs　to　be　considered　in　the　T-peel　test.