Precast　bridge　deck　panels　(PBDPs)　have　been　widely　used　in　bridge　engineering　in　the　past　decade.　However,　as　the　crucial　part　of　connecting　adjacent　precast　members,　wet　joints　are　often　regarded　as　the　weakest　link　in　the　bridge　superstructure.　To　narrow　the　width　of　wet　joints　and　reduce　the　assembling　accuracy　of　U-bars,　a　total　of　five　full-scale　bridge　deck　panels　(BDPs)　were　designed,　which　included　four　precast　bridge　deck　panels　with　wet　joints　and　one　monolithic　cast-in-place　bridge　deck　panel　(MCIPBDP).　The　major　design　variables　included　reinforcement　types　(straight　bar　and　U-bar),　U-bar　lap　details　(non-contact　lap　and　contact　lap)　and　wet　joint　filling　materials　(NC　and　UHPC).　The　axial　tensile　loading　tests　were　conducted　to　evaluate　the　tensile　behavior　of　five　BDPs.　The　experimental　results　demonstrated　that　all　specimens　experienced　ductile　failure,　and　finally　failed　due　to　the　yielding　of　reinforcements.　However,　due　to　the　insufficient　lap　length　of　150　mm　for　the　specimen　P-NS-U200,　bond　slip　appeared,　and　its　ultimate　tensile　capacity　decreased　by　23.2%　compared　to　the　specimen　P-NU-U200.　Regarding　the　specimens　P-NU-U200　and　P-CU-U200,　the　difference　in　ultimate　strength　between　non-contact　lap　and　contact　lap　was　within　3%.　In　addition,　compared　with　the　500-mm-width　normal　concrete　(NC)　wet　joint,　the　tensile　capacity　of　the　200-mm-width　UHPC　wet　joint　was　increased　by　8.8%.　Finally,　based　on　the　strut-and-tie　model　and　failure　mode,　a　modified　calculation　formula　was　proposed　for　estimating　the　tensile　capacity　of　PBDPs.　The　predicted　tensile　capacity　agrees　well　with　the　experimental　results.　The　research　provides　a　reference　for　engineers　to　guide　the　project＇s　construction　and　apply　UHPC　wet　joints　in　practical　engineering.　©　2023　Elsevier　Ltd