The genotype‐environmental interaction of test weight and its components in Eastern soft winter wheat, Triticum aestivum L. em ‘Thell,’ was studied at seven locations in 2 years. Significant effects of location and years for test weight indicate that breeders should include a sample of years and locations when selecting for this character.By removing the differential effects of density, an environmental characteristic, from test weight and dealing only with packing efficiency, a varietal characteristic, a definite gain in genetic variance was obtained.Although test weight and kernel weight under different environments were highly correlated, a covariance analysis demonstrated no genetic correlation between them. This suggests that there are no genes in common controlling test weight and kernel weight in the varieties studied.Utilizing a combined regression analysis to examine genotype stability of test weight, cultivars and advanced breeding lines that had a significant interaction with the nursery means over different environments were identified. These lines were relatively indifferent to environmental variation and always had good test‐weight performance. The analysis also identified genotypes with average sensitivity and lines with extreme or undesirable sensitivity.