Cotton is the leading fiber crop in Pakistan. Its production is negatively affected by living (biotic) and non-living (abiotic) factors and among those, high temperature is most alarming one. Present study was conducted to explore the genetic basis of heat tolerance in upland cotton. Fifty upland cotton genotypes were screened against high temperature on the basis of CMT%, canopy temperature, seed cotton yield, node number of first fruiting branch, days to first effective boll, heat index (HI) and heat tolerance index (HTI).Significant variation was observed among fifty genotypes on the basis of traits used for screening. Seven heat tolerant (CIM-602, Cyto-178, CIM-616, FH-113, CIM-600, FH-114 and KZ-189) and 5 heat sensitive (AA-802, ARK-3, AGC-501, KZ-191 and VH-389) genotypes were selected for crossing. Selected parents were crossed in North Carolina II mating design fashion in the glasshouse. Seven parent and thirty five F1 crosses were field planted under normal and heat stress condition. The analysis of North Carolina Design II revealed that significant genotypic variations exist in all characters under study like plant height, number of sympodial branches per plant, number of monopodial branches per plant, number of bolls per plant, boll weight, GOT%, seed cotton yield, fiber length, fiber strength, fiber fineness, CMT%, node number of first fruiting branch and days to first effective boll. High differences were observed between phenotypic and genotypic variances for various traits which depicted high environmental effect on all studied traits. Among testers AGC-501, AA-802 and VH-389 were found best male combiners for most of the traits. FH-114 showed good GCA effects for plant height and bolls per plant, Cyto-178 for monopodial branches per plant and boll weight, FH-113 for most of the traits studied both under normal and stressed conditions. The crosses FH-114 × ARK-3 and FH-114× AGC-501 showed good SCA effects for most of the traits. Selecting such crosses would be beneficial for the development of superior hybrid through heterosis breeding. Crosses that showed significantbetter parent heterosis for different traits were FH-114 × ARK-3 for plant height,Cyto-178 × AA-802 for monopodia,CIM-616 × ARK-3 for sympodial branches, FH-114 × VH-389 for number of bolls, FH-113 × AA-802 for boll weight, Cyto-178 × KZ-191 for seed cotton yield, CIM-616 × KZ-191 showed maximum fiber length, FH-114 × ARK-3 for fiber strength, FH-114 x KZ-191 for fiber fineness, Cyto-178 × AGC-501 for GOT%, CIM-616 x AA-802 for canopy temperature, FH-114 × VH-389 for node number of first fruiting branch, FH-114 x AA-802 for days to 1st effective boll and CIM-616 × VH-389 for CMT %.As most of the traits were governed by non-additive genes with low narrow sense heritability, so it is suggested that selection may be delayed, or development of hybrid variety would be more effective under high temperature areas.