Wheat (Triticum aestivum L.) is the most widely cultivated crop worldwide and faces a wide range of stresses. To make effective crop improvement decisions, environmental characterization is of paramount importance. This study presents a new methodology for characterizing the environment that enables replacing the conventional arbitrary classification of the environment by a series of environmental covariates that capture and describe the stresses the plant encounters. Three CIMMYT bread wheat populations, combining complementary heat and drought adaptive traits, were grown over 3 yr in northwestern Mexico under limited irrigation, heat stress, and irrigated conditions. The network comprised 15 trials representing seven treatment x year combinations as experimental environments, referred to here as the "Environment". Environmental characterization was performed at the trial scale. Twelve stress thresholds related to eight environmental factors were combined to obtain 11 potential growth limiting factors. Thirty-three environmental covariates were obtained by calculating when these limiting factors occurred for each of three key-developmental-phases across all trials. Cluster analysis allowed grouping environmental covariates into six distinct clusters corresponding to six "environmental scenarios". One representative environmental covariate was extracted from each cluster and taken together explained more than 90% of the variance for yield in the Environment. Principal component analysis discriminated the seven experimental environments and identified its stress characteristics. We conclude that the method developed characterized the main stresses and their impact on average population performance, and the representative covariates efficiently replaced the Environment. As such, they will facilitate the dissection of genotype x environment interaction (GEI) for yield-related traits.