The purpose of this study was to determine the effect of heat stress on carcass weight in a commercial pig population. Phenotypes included hot carcass weight records on 227,043 commercial pigs collected from 2 packing plants located in the states of North Carolina and Missouri; the pigs were a cross between purebred Duroc sires and F1 Landrace x Large White dams. The pedigree file included 553,442 animals, and 60k SNP genotypes were available for 8,232 sires. Weather information was collected from airport weather stations within 100 miles distance from the finishing farms. Average temperature humidity index (THI) was calculated for a period of 70 d before each HCW was collected. The THI measurement was classified as heat stress (THI > 78, n = 32,783) or comfort conditions (THI = < 78, n = 194,260). Analyses were done with an animal model as either a single-trait or 2-trait model using records identified as heat stress and comfort conditions treated as separate traits. Variance components were estimated with AIREML, and traditional and genomic (G) EBV were computed either with BLUP or single-step genomic BLUP (ssGBLUP). Validations were computed for 94 animals from the last generation using the forward prediction method, and reliability of (G)EBV was calculated as R² of predictions based on the training population (all except the last generation) on progeny yield deviations of the last generation. The heritability estimate for hot carcass weight in the single-trait model was 0.20. In the multiple trait model, the heritability estimate was 0.20 under comfort conditions and 0.25 under heat stress, with a genetic correlation of 0.62. Under comfort conditions and heat stress, reliabilities in traditional EBV were 0.22 and 0.14 whereas reliabilities in GEBV were 0.38 and 0.19, respectively. The heritability of carcass weight is higher under heat stress. Use of ssGBLUP increases reliabilities of carcass weight under both heat stress and comfort conditions. Effects of seasonality on carcass weight can be mitigated by genetic selection, especially with the genomic information.