The results of two trials established to study the rate of natural cross pollination in beans (Phaseolus vulgaris L.), in two different growing seasons in Campinas are here discussed. Two cultivares were chosen one having white flowera ('Chumbinho opaco') and tlie other homozygous for the dominant allele for violet flowera ('Mulatinho 1-208'). Two different designs were used but in both enses only the seeds of the white flowered plants were collected and the hybrids identified after sowing in the grenhouse. In the first trial planted in the dry period (February 1959) 10 lines of the white flowered ('Chumbmho opaco') were interplanted with 'Mulatinho 1.208'. It was observed that 12 out 1898 plants had violet flowers, which indicated a rate of cross fertilization of 1.26%. For the trial established in the raining season (September 1961) the white flowered cultivar was planted in the center of a hexagon and in its vértices it was sown seeds of the violet flowered cultivar ('Mulatinho 1-208'). The classification of the seedlings obtained from seeds of 'Chumbinho opaco', indicated that 9 out 677 plants had violet flowers (1.33%). It was noticed a perfect correlation between the flower color and the pigmentation of the cotyledon. This fact will simplify the classification since the hybrids can be identified at seedling stage. Although the rate of natural cross pollination, in Campinas, is relatively low, it is sufficient to permit some crossing between individuals of the same population, increasing slightly the genetic variability. The cultivara isofted obtoined from individuals in the fields and selected by the pedigree method may have appeared as result of such sporadic crossings. It also suggests that in the seed multiplication plots the cultivars should be planted isolated or with borders to avoid possible contamination by natural crossings.