Digestion, growth and laboratory studies evaluated the replacement value of the tropical tree legume Gliricidia sepium for a concentrate feed in kinggrass (Pennisetum purpureum) based diets for growing goats. In the growth trial, 24 intact male goats (10-18 kg) were allotted to three treatments in a randomized design. Nine additional goats used to study digestibility and N balance of the diets were allotted to the treatments in a randomized complete block design with two periods giving six observations per treatment. The (positive) control diet contained 40% concentrate, and gliricidia replaced 0, 50 or 100% of the concentrate, yielding diets containing 0% (G0), 20% (G20) and 40% (G40) gliricidia. Gliricidia was greater in total N concentration than kinggrass, and a greater percentage of the N in gliricidia was potentially ruminally degradable than that in kinggrass. The ruminally degradable N fraction in gliricidia was degraded in situ at a faster rate than that in kinggrass, resulting in a greater estimated pool of N available for rumen function. In the digestion trial, apparent dietary organic matter (OM) and fiber digestibilities, dietary N retained and absorbed, and N utilization efficiency declined in a linear (P < 0.05) manner with increasing gliricidia level in the diet. In the growth trial, OM intake was similar (P = 0.64) among treatments. Differences in daily gain and gain feed ratio were greater (P < 0.06) between the G0 (110 g, 0.19) and G20 (75 g, 0.13) diets than between the G20 and G40 (75 g, 0.13) diets. Lower animal performance on gliricidia-supplemented diets was likely due to lower available dietary energy resulting in lower N digestion and assimilation. All goats on gliricidia-supplemented diets gained weight and maintained a positive N balance, suggesting that gliricidia may replace some of the concentrate used in kinggrass-based diets for growing goats. High costs of grain imports may economically favor lower rates of gain in tropical areas of the world.