This research was performed with the aim yo evaluate the yield and seed quality of soybean genotypes harvested at different dates. Two experiments were conducted, one in the field and the other in laboratory. The field experiment was carried out to obtain seeds and evaluate the yield, whose treatments consisted of ten soybean genotypes [Doko, IAC 11, BR 9 (Savana), CRI, CR3, FT Cristalina, IAC 8, UFV 15 (Uberlândia), FT 11 (Alvorada) and MGBR 22 (Garimpo)], harvested in two different dates, one when physiological maturity was reached and the other 30 days overdue harvest maturity. The experiment was a randomized complet block in a split-plot design with four replications. The genotypes were main treatments and the harvest periods were the subtreatments. In laboratory, the experiment was performed to evaluate the quality of seeds, and consisted of a randomized complete block design with four replications. The treatments were arranged in a fatorial design, in a combination of ten soybean genotypes and two harvest dates. In the field conditions, the yield and seedling emergence were evaluated. In the laboratory, germination and vigor of the seeds were evaluated by means of the germination test, the conductivity test, and the accelerated ageing test. The quality was also evaluated by a sanity test. It was conclued that: 1) there were significant differences in yield among genotypes, with the highest and lowest productivity observed respectivelly for FT 11 and CR1, either when harvested at physiological maturity or 30 days after harvest maturity; 2) the delayed harvest impaired productivity, as well as germination and vigor of the seeds for alI genotypes, and it caused an increase in the incidence of fungus as well; 3) the seedling emergence was similar for all genotypes when harvested at physilogical maturity, however, there were greater significant differences between the genotypes at the retarded havesting; 4) the results of the different tests identified the genotypes Doko and UFV 15 as the best physiological seed quality; 5) FT 11 and Garimpo presented seed quality relatively good when harvested in physiological maturity, but they were the ones which lost the most capacity of germination when harvest was delayed, showing that seed producers shall not delay their harvest.