The study was conducted to evaluate the suitability of Leucaena leucocephala for fuelwood and pole production using agroforestry in a semiarid environment. The trial site, planted in February 1980 at Mafiga, Morogoro, was a split plot design with four main plots: (1) Leucaena intercropped with maize; (2) Leucaena intercropped with beans; (3) Leucaena monoculture and clean weeded; and (4) Leucaena monoculture and spot weeded. Each main plot included three tree-spacing sub- plots, 3 X 3 m, 4 X 4 m and 5 × 5 m, and replicated four times. Each of the subplots comprised at least 25 trees. Twenty-four trees representing all the diameter classes in the trial were harvested in March 1984 and used to determine mean plot height, volume and biomass by regression analysis. Nutrient content was determined using routine laboratory methods. Mean subplot height varied from 7.0 to 7.8 m, whereas diameter at the stem base varied from 10.2 to 14.9 cm. The volume varied from 10.3 to 32.8 m 3 ha -~ and total biomass values from 13.5 to 31.1 metric t ha -1. Trees under spot weeding showed the poorest growth, whereas those under clean weeding showed the best. Spacing influenced both diameter and height growth, with the widest spacing producing the largest trees. Because of a higher tree population, however, the lowest spacing showed the highest volume and biomass production. Intercropping with maize and beans showed significantly higher volume and biomass production as compared to spot weeding. However, there was no significant difference between trees under clean weeded treatments and those under intercropping. Nutrient concentra- tions were high and comparable to other pasture legumes. The total nutrient accumulation in the aerial biomass in kg ha- 1 was 201, 25.1,267, 106, 34 and 18 for nitrogen, phosphorus, potassium, calcium, magnesium and sodium, respectively.