Forty two, 21 males and 21 females, weaned Sudan Desert lambs were divided according to sex and live weight into three equal groups (3 x 14). Each group was randomly fed ad libitum for 11 weeks on one of three diets : humra Aristida funiculata Trin. & Rupr., humra plus limited amound of cotton seed cake, and a balanced ration. After those 11 weeks of differential feeding, each group was fed ad libitum on the balanced ration for 12 weeks. Six lambs (3 males and 3 females, 2 from each group), were randomly slaughtered at the start of the differential feeding period. Half of the remaining lambs (3 x 6) in each treatment group (3 x 12) were slaughtered at the end of the differential feeding period and the other half (3 x 6) at the end of the experiment. Live-weight growth of lambs fed humra, with or without supplemental cottonseed cake, was restricted compared to lambs fed the balanced ration. Upon refeeding, previously restricted growing lambs expressed compensatory growth. The pattern of relative increase in weight of body components was not affected by nutritional history but was influenced by body weight. Following refeeding, animals fed earlier humra with or without cottonseed cake, recovered in body conformation and composition. It was concluded that, supplementation of range lambs with a source of protein during the season of graze scarcity and afterwards feeding them on high energy yielding rations shortly before slaughter is recommended under the conditions of the Sudan.