The diploid species of Coffea so far investigated, are self sterile, while tetraploid C. arabica, besides being self-fertile, is an almost completely autogamous species. Natural crossing in this species is promoted either by wind or by insects while gravity probably only plays a limited role. The trial here discussed was established to study the effect of certain native and also of the European honey bee in the promotion of the pollination in C. arabica var. bourbon, and also to test their influence on the yield of coffee. As it was difficult to specifically determine the effects of only this group of insects, the data represent the total effect on coffee yield and seed characteristics, of the exclusion of all pollinating insects. A group of 135 coffee "hills", each with four seedlings, and planted in 15 rows of 9 "hills" each, was used for this purpose. A row consisted of two plots of three "hills" each, besides the borders. One "hill" in each plot was always covered with muslin cloth as shown in figure 1. The trial comprised 26 replications. About 20 meters away from the trial, 15 beehives were established, basically of the following species: Nannotrigona (Scaptotrigona) sp., Tetragona (Tetragonisca) jaty, Melipona quadrifasciata anthidioides, Plebeia sp. and Nannotrigona (Na.) testaceicomis. In the course of the experiment, the species Melipona Scutellaria and Cephalotrigona capitata were added. There was an apiary of honey bees (A. mellifera) at about 700 meters from the experiment. The muslin protections to exclude insects were palced one day before the opening of the flowers and removed, some 10 to 15 days later, when all flowers bad dried corolla?. These treatments were repeated every year from 1951 to 1956, in order to collect information on the yield of each of the "hills" and also on the seed types produced and their size. During 1953, 1954 and 1955, data were also collected concerning the percentage of fruit set and seed types based on a certain number of previously counted flowers. It was noticed that the nectar of the coffee flowers was still abundant even 24 hours after anthesis; its sugar content was about 38%. Apis mellifera was the species most frequently found visiting the Bourbon flowers. Among the native bees, the Plebeia sp. and Nannotrigona (Na.) testaceicomis were found in higher numbers, collecting mainly pollen. During rainy weather, nearly half of the A. mellifera bees present was seen collecting only pollen. On sunny days a large number of bees was observed visiting the coffee flowers, A. mellifera being found in larger numbers at about 12 a.m., at temperatures of 30 to 32°C. The Plebeia sp. and N. (Na.) testaceicornis start their flights later than A. mellifera. Of several A mellifera bees marked, only one was seen four times during 24 hours in an area of 13m x 6m. The inconstancy of these bees may be due to the short duration of the coffee flowering period. According to the data obtained, large bees, as A. mellifera and Melipona quadrifasciata were more efficient to promote pollination than the small ones. Plebeia sp., Tetragonisca (Te.) jaty and N. (Na.) testaceicornis were found on older flowers collecting nectar from the base of the displaced and almost dried corolla tubes. The meliponin bees Plebeia sp., N. (Na.) testaceicomis, Tr. (Tr.) ruficrus were sometimes frequently seen taking nectar through the slit like holes made near the base of corolla tubes by Tr. (Tr.) hyalinata and by the wasp Brachygastra augustii. Additional observation indicated that A. mellifera and the native bees M. quadrifasciata, Cephalotrigona capitata prefer the larger and more scented flowers of Coffea Dewevrei hybrids, found farther away from lhe Bourbon trial. N. (Scaptotrigona) sp. although present in strong colonies, was seen only once on coffee flowers. In spite of the large number of insects collecting nectar and pollen on Bourbon coffee flowers, the yield data here presented do not show significant differences between the protected trees in comparison with the control, without any protection. In these plants only a certain tendency towards a higher production was observed, during four of the six consecutive years. This effect may be the result of an intensified liberation of pollen, simply due to insect contacts,the pollen then being transported by wind. The somewhat lower yield of the protected plants may also be the result of the reduction of light due to the use of muslin protection which remained on the trees for about 10 to 15 days after each coffee blossoming, 3 to 4 times every year. The data on fruit setting, also indicate a higher percentage in the unprotected trees. The percentage of peaberry (²) seeds was significantly higher in the covered plants. The percentage of "elephant" (³) seeds was also higher, but not significantly, in this same group of plants. The causes of these abnormalities are not yet well understood. There were no differences in the amount of empty seeds and in the size of the beans in both groups of plants. The influence of bees on coffee pollination must be restricted nearly only to the diploid self-sterile species of Coffea. The results here presented for C. arabica var. bourbon indicate that the role played by insects in promoting pollination and increasing the yield, seem to be of secondary importance for this variety of coffee.