Physiological breeding involves application of practical phenotyping tools to dissect the genotype by environment interaction of genetic resources developed for target environments. A subset of the WAMI population (140 lines) was evaluated for terminal heat tolerance at two locations Faisalabad (temperate) and Multan (heat stressed) in Pakistan during 2017-18. Phenotyping was performed for canopy temperature (CT) during grain filling, normalized difference vegetation index (NDVI) and SPAD chlorophyll at anthesis+10 days. Plant height, days to heading and maturity including 100-seed weight, seed yield (SY), biomass yield (BY), number of fertile tillers (NFT, m-2) and harvest index (HI) were recorded. Multivariate analysis was performed to explore the heat tolerant lines and association between CT, NDVI, SPADChl, NFT, biomass yield (BY), seed yield (SY) and harvest index (HI). Data of 140 lines were subjected to principal component analysis (PCA). Results of PCA were examined with biplots developed from PC 1 for temperate and PC 2 for heat stressed environment using Origin Pro 9.1 software (OriginLab Corporation, Northampton, USA). The PC 1 and PC 2 showed more total variability (64.10 %) under heat stressed than temperate environments (59.50%). Associations between physiological and yield traits were explored with the vectors angles. The biplots showed strong positive association among yield components (NFT, BY and SY) while the association of physiological traits SPAD-Chl, CT and NDVI were also positively associated with yield R2=0.22, R2=0.24 and R2=0.13 (p=0.05), respectively. Similarly, the genotypes with better performance for specific traits were closely associated with respective trait vectors. Wheat genotypes (41, 111) with high seed yield, cooler CT and high expression of SPAD-Chl can be used as promising heat adaptive sources in physiological breeding for future climate change.