Generally,　the　cohesive　healing　within　the　asphalt　composites　and　the　adhesive　healing　from　the　asphalt-aggregate　interface　always　occur　simultaneously　in　the　asphalt　concrete　with　growing　damage.　This　study　aims　to　investigate　the　cohesive　and　adhesive　healing　behaviors　of　asphalt　binder　respectively　by　utilizing　the　newly　developed　linear　amplitude　sweep　based　healing　(LASH)　and　binder　bond　strength　based　healing　(BBSH)　tests.　Seven　unmodified　asphalt　binders　and　one　styrene–butadienestyrene　(SBS)　polymer　modified　binder　are　selected　in　this　study.　A　fracture　mechanics　based　crack　growth　approach　is　proposed　to　define　the　cohesive　healing　recovery　in　the　LASH　test　with　two　percent　healing　parameters,　in　terms　of　the　%Ha　of　the　crack　length　recovery　and　the　%HA　of　the　sample　intact　area　recovery.　In　pre-failure　conditions,　the　measured　%Ha　can　at　least　reach　beyond　30%　while　the　%HA　is　normally　within　30%　for　tested　binders.　Meanwhile,　the　%Ha　results　show　more　consistent　trends　to　the　traditional　simplified-viscoelastic　continuum　damage　(S-VECD)　based　LASH　data　interpretation.　Furthermore,　the　cohesive　failure　occurrence　is　verified　to　significantly　impact　the　healing　performance　comparison.　The　cohesive　healing　behaviors　of　neat　and　SBS　binders　respectively　in　pre-failure　and　post-failure　conditions　are　distinguished　to　each　other,　which　are　validated　either　from　S-VECD　or　crack　growth　approaches.　Regarding　to　the　binder　adhesive　healing　behavior,　it　is　observed　that　the　presence　of　water　and　the　using　of　granite　substrate　generally　have　positive　effects　on　adhesive　healing　recovery　in　the　BBSH　tests.　The　SBS　modification　generally　displays　lower　adhesive　healing　percentages　than　the　neat　binders　especially　in　the　wet　conditions.　Though　no　clear　relationships　can　be　found　between　the　cohesive　and　adhesive　healing　behaviors　in　this　study,　it　is　recommended　to　quantify　the　binder　contribution　to　the　adhesive　healing　process　in　the　future　work.　©　2021　Elsevier　Ltd