Studies involving several accessions of Sesbania sesban were undertaken in the laboratory, nursery and field. The laboratory study was carried out at Iowa State University to determine the best pretreatment methods for sesban's seeds, based on four provenances. There were varying responses among the provenances to the different pretreatment methods. A 1-hour soak in hot water of an initial 100°C was recommended because it had the best general effect;At the Moi University Nursery, Eldoret, Kenya, an inoculation study was undertaken to determine symbiotic effectiveness among 12 accessions of S. sesban. Effects were measured in height growth, leaf color, nodule number, shoot fresh and dry weights, and root fresh and dry weights. Accession variation was statistically significant with all the traits. Symbiotic effectiveness ranged from 11.4-60%;Field research was carried out at six different sites in Kenya: Eldoret, Turbo, Ratta, Got Ramogi, Katumani and Muguga. Fuelwood characteristics were measured 8 months after planting. The traits measured/estimated included height, basal diameter, fuelwood volume, wet wood density, dry wood density and moisture content. Volume production was greatest among Kenyan accessions and an Ethiopian accession with the highest volume per tree over all sites being 1613 cm[superscript]3. Specific gravity ranged from 0.26-0.54 with sites combined;Length of live crown, number of branches, branch length, foliage abundance, height and basal diameter were measured at 2, 5 and 8 months after planting to determine age effects on selection. Age x accession effects were not significant, whereas, large accession x site interaction was evident;Estimates for heritability ranged from 0.23-0.88 for the different traits within sites and across sites and ages. Katumani and Muguga recorded lower estimates, thereby indicating the presence of larger environmental variability. Stability analysis revealed that the accessions studied were generally unstable in their response to the environments. Because of the high accession x site interaction, broadly adapted genotypes may not be the best approach to sesban breeding. As a result of large genetic differences, moderately high heritability estimates, fast growth, and early flowering in sesban, it is recommended that selection should be done early in a breeding program.