Future climate projections and impact assessments are critical in evaluating the potential impacts of climate change and climate variability on crop production. Climate change impact assessment in combination with crop, climate models under different climate change scenarios is uncertain and it is challenging to select an appropriate climate scenario. This study quantifies the uncertainty associated with projected climate change impacts on cotton yield in Punjab, Pakistan using 29 general circulation models (GCMs) under high and moderate representative concentration pathway (RCP) scenarios (4.5 and 8.5) at near-term (2010–2039) and mid-century (2040–2069) time spans. Cropping System Model (CSM) CROPGRO-Cotton (DSSAT v 4.6) was calibrated and evaluated with field experiment data collected under arid/semi-arid climatic conditions. Enormous variation was observed in GCMs climatic variables, which were therefore classified into different categories. According to mean ensemble of 29 GCMs, there is a projected increase in seasonal average temperature 1.52 °C and 2.60 °C in RCP 4.5 and 1.57 °C and 3.37 °C in RCP 8.5 scenario as compared to the seasonal baseline (31.48 °C) in near-term (2010–2039) and mid-century (2040–2069), respectively. Maximum consensus by GCMs revealed the increase in temperature of 1.2–1.8 °C and 2.2 to 3.1 °C in RCP 4.5 scenario while 1.4–2.2 °C and 3.0–3.9 °C increase is expected under RCP 8.5 for near term and mid-century time periods, respectively. Similarly, rainfall changes are expected −8% to 15% and −5 to 17% in RCP 4.5 scenario while −8 to 22% and −2 to 20% change is expected under RCP 8.5 scenario in near term and mid-century time periods, respectively. Seed cotton yield (SCY) are projected to decrease by 8% on average by 2039 and 20% by 2069under the RCP 4.5 scenario relative to the baseline (1980–2010). Mean seed cotton yield is projected to decrease by 12% and 30% on average under the RCP 8.5 scenario. Uncertainties were observed in GCMs projections and RCPs due to variations in climatic variables projections. GCMs, GFDL-ESM2M (45% and 35%), GFDL-ESM2G (28% and 43%) and MIROC-ESM (39% and 70%) predicted the higher mean SCY reduction ensemble of cultivars than others under emission scenario of 4.5 in near term and mid-century, respectively. Lower SCY reduction was revealed in CCSM4, HADGEM2-CC, HADGEM2-ES, INMCM4 and CNRM-CM5 due to mild behavior of climatic variables especially temperature under RCP 4.5 in the near-term and mid-century. High reduction in mean SCY (16%–19%) is expected in CMCC-CMS, IPSL-CM5B-LR, GISS-E2-H, GFDL-ESM2M and GFDL-ESM2G under the RCP 8.5 scenario. However, under the same scenario, mean SCY increases by 1% in HADGEM2-ES and by 4% in HADGEM2-CC relative to the baseline yield (4147 kg ha−1). GFDL-ESM2M and GFDL-ESM2G are hot and dry while HADGEM2-ES and HADGEM2-CC are hot but wet, resulting in less cotton yield loss. MIROC-ESM and GFDL-ESM2G projected a severe reduction in mean SCY (70% and 69%) due to a steep increase in maximum and minimum temperature (6.97 °C and 4.38 °C, 4.91 °C and 3.70 °C), respectively and sever reduction in rainfall by mid-century and may call worse case scenarios. Climate models like, CCSM4, HadGEM2-CC, HadGEM2-ES, INMCM4, CanESM2, CNRM-CM5, ACCESS1-0, BNU-ESM and MIROC5 are found less uncertain and showed stable behavior. Therefore, these models can be used for climate change impact assessment for other crops in the region. Adaptation management options like five weeks early sowing than current (10-May), increasing nitrogen fertilization (30%), higher planting density (18% for spreading and 30% for erect type cultivars) and 17% enhanced genetic potential of cultivars would compensate the negative impacts of climate change on cotton crop. This study provide valuable understandings and direction for cotton management options under climate change scenarios. This multi-model and multi-scenario analysis provides a first overview of projected changes in temperature and precipitation, cotton yield and potential management options under changing climate scenarios in arid to semi-arid climatic conditions of Punjab-Pakistan.