The effects of spray timing and number of benomyl applications on the seed quality was evaluated through a field experiment, in a randomized complete block design. Treatments were: a) check, no fungicide; b) normal, two appiications, the first at the initial pod set and the second two weeks later; c) normal + 1, a third application added to the normal schedule, fourteen days after the last application; d) delayed 1, the normal schedule delayed two weeks; e) delayed four weeks. Greater leaf retention and better control of foliar diseases was observed in treatments c, b and d in order. Treatment e did not difter from the heck. The 250 seed weight was significantly increased by all fungicide treatments, however better results were obtained in treatment c. Treatments b, d and e were not different from each other. The fungicidal applications resulted in significant increases in vigor but the increases in the germination percentage were not significant. The fungi bioassay tests revealed that Diaporthe phaseolorum var. sojae was the predominant organism within seeds. Followed by Cercospora kikuchii (8.4%), Aspergillus sp. (2.8%), Fusarium sp. (1.4%) and Miscellaneous (6.8%). The levels of infected seeds in the fungicidal treatments were lower but they were not statistically different from the control.