The floral biology of Sesbania sesban ( L.) Men'. ( Leguminosae: Papilionoideae) was studied inrngeneral, and in particular the level of out-crossing was detennined under conditions of openpollination.rnFloral biology studies were conducted on one cultivated population (ILRI AccessionrnNo. 15022) of S. sesban at the compound of the International Livestock Research Institute (ILRI)rnin Addis Ababa. For the purpose of estimating out-crossing rate, seeds were collected from S.rnsesban populations at Addis Ababa (Ethiopa), Debrezeit (Ethiopia) and Zambia.rnStigma receptivity and anther dehiscence time vary considerably for different flower bud sizes.rnStigma become receptive before shedding of pollen grains and lor maturity of pollen grains.rnTherefore, the flowers of this species were detennined to be protogynous. There were significantrndifferences between mean bud size (mm) with receptive stigma (x = 14.1±1.46) and those withrnnon-receptive ones (x= 12.1±1.84). Also between mean size of buds with dehiscing anthers (x =rn15.39 ± 2.28) and those with non-dehiscing anthers (x =10.68 ± 2.36) at P < 0.001 level ofrnsignificance. Further more, pollen tubes were found to be abundant in more mature flower buds (xrn= 16.53 mm) from deliberate tripping. Gennination of self-pollen under natural condition takesrnplace at a very late bud stage i.e. when a flower bud size of 22 rum and above is attained. On thernother hand, tripping of flowers indicated that pollen gennination could take place at an early budrnstage (stage 2, x = 14 rum). Pollen tube growth was more abundant in those styles, which receivedrnhand tripping treatment than with forceps or toothpick tripping. Seed production (no.l pod) wasrnalso higher due to hand tripping (x = 29.9 ± 1.27) than due to bagging alone (x = 23.67 ± l.84) orrntoothpick tripping (x = 24.11± 2.25), although not statistically significant. However, seedrnproduction from open-pollination (x = 34.11 ± 2.60) was significantly higher than from therntripping and bagging treatments.rnBecause of difficulties in preparing an optimum medium for maximum pollen grain gernlination inrnvitro and due to short hydration time (6 hours of incubation), 49.43% gennination was the averagernscore obtained in 20% sucrose. On the other hand, with fluorochromatic reaction test an average ofrn90.51 % viability was scored from flower buds that fall in the range of 11 to 22 rum. Flower bud sizernwas significantly negatively correlated (r = -0.61, P =0.000) with pollen grain viability and pollenrngrain viability was also found to correlate (r = - 0.39, P = 0.000) inversely with germination. Therernwas a significant positive correlation (r = 0.87, P = 0.000) between flower bud size and genninationrnatP < 0.05.rnAlthough the floral structure of S. sesban could favour self-pollination, the delay in self-pollenrngelmination under natural condition rather seems to favour out -crossing in the presence ofrnappropriate pollinators. Based on the electrophoretic assay of a large number of progenies on onernpolymorphic locus of Malate dehydrogenase (Mdh), on average, an estimated 64% of the seedsrnwere found to be derived from out-crossing, ranging from 0.395 to 1.105 among populations. Therngenotype frequencies of the populations of Accession 15022 and Accession 15019 were in line withrnthose of the Hardy-Weinberg expected genotype proportions, whereas genotype frequencies of thernZam-R024 population deviated from the expectations indicating the occurrence of non-randomrnmating in this population.rnIt is possible to conclude that S. sesban can under go complete self-fertilization in the absencernpollinators and also complete out-crossing in the presence of suitable pollinators.