Increasing scarcity of irrigation water is a major threat to sustainable production of cotton (Gossypium hirsutum L.). Identifying genomic regions contributing to abiotic stress tolerance will help develop cotton cultivars suitable for water-limited regions through molecular marker-assisted breeding. A molecular mapping F2 population was derived from an intraspecific cross of the drought sensitive G. hirsutum cv. FH-901 and drought tolerant G. hirsutum cv. RH-510. Field data were recorded on physiological traits (osmotic potential and osmotic adjustment); yield and its component traits (seedcotton yield, number of bolls/plant and boll weight); and plant architecture traits (plant height and number of nodes per plant) for F2, F2:3 and F2:4 generations under well-watered versus water-limited growth conditions. The two parents were surveyed for polymorphism using 6500 SSR primer pairs. Joinmap3.0 software was used to construct linkage map with 64 polymorphic markers and it resulted into 35 markers mapped on 12 linkage groups. QTL analysis was performed by composite interval mapping (CIM) using QTL Cartographer2.5 software. In total, 7 QTLs (osmotic potential 2, osmotic adjustment 1, seedcotton yield 1, number of bolls/plant 1, boll weight 1 and plant height 1) were identified. There were three QTLs (qtlOP-2, qtlOA-1, and qtlPH-1) detected only in water-limited conditions. Two QTLs (qtlSC-1 and qtlBW-1) were detected for relative values. Two QTLs (qtlOP-1 and qtlBN-1) were detected for well-watered treatment. Significant QTLs detected in this study can be employed in MAS for molecular breeding programs aiming at developing drought tolerant cotton cultivars.