U nderstanding and quantifying photoperiod × temperature interactions often directly aff ects soybean [Glycine max (L.) Merr.] breeders and producers when selecting varieties, determining dates of planting, predicting dates of fl owering and maturity, and predicting fi nal yields (Zhang et al., 2001). Eff ect of the photoperiod response on area of adaptation is more pronounced in the soybean than in any other major crop. As soybean is classifi ed as a short-day plant, sensitivity to photoperiod is a hindering factor in increasing its adaptation range. When soybeans are cultivated under short-day conditions, in out-of-season plantings or in low latitude, those plants with the classic response to photoperiod fl ower early and result in short plants and low grain yields (Carpentieri-Pípolo et al., 2000). The length of the growing season for photoperiodic sensitive crops such as soybean is defi ned by complex interactions between temperature and photoperiod (Raper and Kramer, 1987). ABSTRACTMaturity classifi cation is an important concept to provide the best allocation of resources for soybean [Glycine max (L.) Merr.] research and commercialization. A similar maturity group system used in North America is being used for some seed companies in Brazil and needs research to improve its use. This study evaluated the maturity stability of 48 midwestern and 40 southern Brazilian commercial cultivars ranging from North American maturity groups VI to VIII at 15 locations. Relative maturity groups were attributed to all cultivars. All trials were planted in the fi rst half of November. The effect of location was very important in infl uencing the number of days to maturity, number of days to fl owering and reproductive growth period (RGP). The genotype × environment interaction, although statistically signifi cant, was much lower than the individual effects of environment and genotype for all traits and regions. Genotype × latitude and genotype × altitude, considering also years of evaluation, were generally low or nonsignifi cant. A recommended list was developed of the most stable genotypes and, consequently, of the most suitable check genotypes for each maturity group classifi cation in the southern and midwestern regions. Results indicate that the use in Brazil of a maturity group system similar to that used in North America to classify soybean genotypes is an effi cient method for describing relative maturity on a broad environmental basis.