Twenty four similar trials with corn were conducted throughout Central America during 1989 and 1990. The specific objectives were: i) to evaluate the effect of intercropping legumes with corn, ii) to establish P requirements in a system of intercropping corn with legumes, iii) to evaluate soil coverage by the legume as an indicator of a potential for erosion reduction and iv) evaluate weéd control in the intercropped system. The following legumes were tested: Stizolobium deeringianum Bort (mucuna), Vigna unguiculata L. and Canavalia ensiformis L. each one with P doses of 20 and 40 kg/ha. In general, the results obtained in 1989 were similar to those of 1990. The average corn yield for the different locations varied between 1.8 and 6.0 t/ha in 1989 and 1.0 and 5.6 t/ha in 1990, indicating the great environmental variability of the corn zones of Central America. The corn yield was always lower with legumes than without them, at the same rate of P to both years, C. ensiformis reduced the corn yield by 0.310 (±0.32) t/ha, V. unguiculata by 0.540 (±0.48) t/ha and S. deeringianum by 0.610 (±0.69) t/ha. The response to P was variable and depended on the year, the location and the legume. A positive response to P applied to corn without the legume was observed in both years. The was a positive response to P in corn intercropped with.§, deeringianumin 1982 in corn intercropped and with C. ensiformis in 1990. The legume evaluation as a mulch (average of 8 trials) indicate that at 30 days after the corn planting, V. unguiculata gave the best coverage with 62%, followed by C. ensiformis, with 38% and S. deeringianum with 35%.