An experiment was conducted in five locations in the San Juan valley, Dominican Republic, to study the response of splitting nitrogen fertilization and time of application on the seed yield of common bean (Phaseolus vulgaris L.). Trials were established between november 5 to 14,1997. In each location nine treatments were tested under a Complete Random Design. There was no response to nitrogen fertilization in a single or splitted application in locations with high nitrogen content in the soil; whereas in location with soils deficient in this nutrient, a significant response (P<0.05) was observed in both, fertilization rate and in the splitting of the doses. There were significant differences among locations (P<0.05) and the interaction location X fertilization treatment was not significative.