Two tropical legumes and a temperate legume were studied during the course of this work. These were Leucaena leucocephala (cv. Peru), Canavalia ensiformis(cv. unknown) and Lupinus albus (cv. Vladimir [Kievskji mutant]) respectively. The dried ground leaf obtained from Leucaena leucocephala (LLM), the seeds of Canavalia ensiformis (jack bean; JB) and Lupinus albus (lupin) were the materials used in the study. Liquid chromatographic methods (HPLC) were developed for the analysis of mimosine and 3-hydroxy-4(lH )-pyridone (3,4-DHP) in LLM, Leucaena seed (LS) and chick excreta. Neither mimosine nor 3,4-DHP wftS detected in the serum of chicks fed LLM. Poor and variable recoveries of mimosine and 3,4-DHP were obtained when these were added to serum The analysis of canavanine in JB and the serum of chicks fed JB and canavanine, was also accomplished using HPLC. Canaline was not detected in any of the samples analysed by HPLC although recovery of added canaline to serum, JB and excreta was high. A small amount of what appeared to be canavanine was detected in lupin. Saponins and tannins were found in all the legumes under study. Trypsin inhibitors were detected in all but the lupins Inclusion of LLM in chick diets reduced their performance. Addition of Fe(III), polyethylene glycol (RAM = 4000) and cholesterol to LLM diets improved chick performance almost to that of chicks fed control diets. Cooking LLM alsoimproved chick performance. Addition of enzymes to LLM did not improvechick performance and did not improve the apparent metabolisable energy (AME) of LLM. Dietary inclusion of mimosine or LS, to supply the same amount of mimosine as that from LLM, did not restrict chick performance to the same extent as LLM Inclusion of autoclaved jack bean (JB) in chick diets caused a severe reduction in chick performance. The reduction in performance was not matched byinclusion of canavanine at the same level as that from JB. The inclusion ofextracted JB also reduced chick performance. Germination of the JB, prior to autoclaving and dietary inclusion, did not reduce canavanine levels nor was chick performance improved. Addition of arginine to JB diets improved performance of chicks but additional lysine had no beneficial effect. Lupin diets perm itted chicks to perform much better than LLM or JB diets. Autoclaving made little difference to the AME of lupins although the results were probably confounded by the presence of M aillard reaction products. Addition of enzymes to lupins increased the concentration of lower molecular weight carbohydrates but only had a small beneficial effect on AME of lupins for chicks.