Existing　literature　has　made　substantial　efforts　to　examine　the　relationships　between　land　surface　temperature　(LST)　and　urban　landscape　patterns　(ULPs).　However,　the　inconsistent　findings　from　studies　on　LST　conducted　in　different　cities　lead　to　concerns　about　the　significance　and　importance　of　ULPs.　Moreover,　insufficient　attention　has　been　paid　to　vertical　ULPs　and　variations　in　their　thermal　effects　over　space　and　time.　This　study　conducts　a　comparative　analysis　in　38　Chinese　megacities　across　different　seasons　at　the　street　block　level　to　identify　regularities　and　differences　in　ULP-LST　linkages　using　geographical　open　data.　The　study　quantifies　ULPs　with　an　amount　of　widely　used　and　new　two-　and　three-dimensional　spatial　metrics　from　three　aspects—city　plan　patterns　(CPPs),　building　patterns　(BPs),　and　land　use　patterns　(LUPs)—based　on　Conzen＇s　townscape　analysis　framework.　Results　reveal　that　the　consideration　of　overall　or　specific　aspects　of　ULPs　can　enhance　the　explanation　of　spatial　variations　in　LST,　particularly　during　summer　and　spring.　The　improvements　are　highest　for　LUPs,　followed　by　CPPs　and　BPs.　Regardless　of　seasons　and　cities,　building　arrangement,　built-up　areas,　greens,　water　bodies,　elevation,　slope,　and　road　density　are　the　most　influential　ULP　indicators,　whereas　block　size,　sky　view　factor,　building　density,　and　building　height　present　limited　or　unintended　effects.　Furthermore,　our　results　indicate　the　time-　and　place-varying　relationships　between　ULPs　and　LST,　and　some　ULP　indicators　demonstrate　two-sided　effects　on　LST　across　different　seasons　or　cities.　We　suggest　that　optimizing　building　layout　and　land　use　composition　to　increase　green-blue　spaces　and　urban　shading　zones　may　be　more　effective　for　alleviating　the　urban　heat　island　effect　than　changing　urban　density.　©　2024　Elsevier　Inc.