Six tropical feedstuffs were analyzed for proximate composition, Van Soest components, mineral contents and amino acid composition. The crude protein contents (dry basis) of cassava tuber meal, sesame oil meal, coconut oil meal, rubber seed meal, cassava leaf meal and sweet potato leaf meal were 2.9, 35.2, 21.8, 12.0, 20.2 and 10.4%, respectively. The corresponding values for acid detergent fiber were 5.0, 23.5, 29.4, 39.1, 35.0 and 46.4%, respectively. Cassava tuber meal contained 88.4% nitrogen-free extract, but was poor in minerals and amino acids. Sesame oil meal had a silica content of 12.2% possibly indicating adulteration with sand; it was rich in all the minerals, especially Ca (2.32%), Mg (.56%), P (1.04%) and Fe (.32%). Coconut oil meal contained high amounts of Na (852 ppm) and K (1.83%), whereas the Zn (249 ppm) and Mn (252 ppm) were higher in cassava leaf meal. With the possible exception of sesame oil meal, the sulphur amino acids are likely to be the most limiting when these feedstuffs are fed to livestock. True metabolizable energy (TME) values of cassava tuber meal, cassava leaf meal and sesame oil meal were determined using 24 adult Single Comb White Leghorn roosters. After 24 hours of fasting, six roosters were force-fed 25 g of each of these feedstuffs. Six roosters were fasted for another 48-hour period and served as negative controls. Excreta were collected for two consecutive 24-hour periods. Length of collection period had no effect on the TME of cassava tuber meal and sesame oil meal, indicating that a 24-hour collection period was adequate for these feedstuffs. A collection period of 48 hours or longer was, however, required for roosters force-fed cassava leaf meal. The mean TME values (dry basis) of cassava tuber meal, cassava leaf meal and sesame oil meal were determined to be 3.76 ± .06, 1.99 ± .28 and 2.42 ± .09 kcal/g, respectively. Three digestion trials, each involving 12 crossbred gilts averaging 35.2 kg body weight, were conducted to determine the effects of crude fiber and Virginiamycin on digesta rate of passate (RP) and nutritional parameters. Two levels (3.2 and 7.3%) of crude fiber and two levels (0 and 11 ppm) of Virginiamycin were used in a 2 x 2 factorial arrangement of treatments. All diets were isonitrogenous and isocaloric. RP was determined by noting the time required for a change in feces color following the addition of 0.5% chromic oxide to the diet. The high fiber diet had a faster (P<.001) RP and this was associated with depressions (P<.001) in the digestibility of dry matter (DM), energy (E), cell content, cell wall (CW), acid detergent fiber (ADF), hemicellulose, cellulose, lignin and ash. Virginiamycin supplementation slowed (P<.01) the RP of both low and high fiber diets, but improved (P<.10) the DM, E and CW digestibility of the high fiber diet only. Non-significant improvements (P<.15) in the digestibility of ADF and cellulose were also observed when Virginiamycin was supplemented to the high fiber diet. Fiber increased (P<.05) fecal nitrogen, whereas Virginiamycin supplementation decreased (P<.05) fecal nitrogen. Results suggest that Virginiamycin supplementation improved the energy utilization in a high fiber diet, but had little effect on a low fiber diet. A study was conducted, concurrent to the first total collection digestion trial, to evaluate the applicability of chromic oxide indicator method to estimate the digestibility of nutrients. Digestion coefficients derived from the indicator method were in close agreement with those determined by the conventional total collection method for the low fiber diet, but underestimate the values for the high fiber diet. / Master of Science