Quantification of the drought tolerance of wheat (Triticum spp.) genotypes is of interest to physiologists investigating traits associated with adaptation to dry conditions and to breeders developing cultivars for dry areas. The objective of this research was to assess drought tolerance in terms of minimal depression of yield in dry compared with favorable environments (drought susceptibility index, S) with a superiority measure (P) based on the sum across environments of the mean‐squared differences between genotype yield and the maximum yield in each environment. Twenty‐five hexaploid (T. aestivum L.) and 16 tetraploid (T. turgidum L. var. durum) wheat genotypes were grown in separate experiments under dry and irrigated conditions on a Swinton loam (Aridic Haploboroll) soil. There was year‐to‐year variation in S within genotypes and changes in genotype ranking within years. No single cause for this variability was identified. The S index did not differentiate between potentially drought‐tolerant genotypes and those that had low yield potential from other causes. P was correlated with mean yield in both the hexaploid (−0.96**) and tetraploid (−0.94**) groups (both significant at P = 0.01). However, P was strongly influenced by high yield environments; this was alleviated by standardization of the yield data. Although P appears to be a possible method to quantify average superiority of genotypes across locations, its suitability requires further assessment. There seems to be no simple technique to quantify drought tolerance that would assist physiologists in choosing genotypes in which to evaluate putative drought tolerance mechanisms.