The research was conducted at the International Livestock Research Institute, Debre Zeit Research Station, which is situated in the central highlands of Ethiopia. Leaves of Lablab purpureus and the multipurpose trees (MPT), namely, Leucaena pallida 14203 and Sesbania sesban 1198 were harvested and dried in a shade at the research station. Seventy individually penned Menz ewes with a mean initial weight of 14.77±0.32 kg (mean±S.D.) were randomly assigned to seven treatments in a randomized block design. The ewes were supplemented daily with three levels of L. pallida 14203, namely, LP1 (8.3 g DM kg −1 LW), LP2 (11 g DM kg −1 LW), LP3 (13.3 g DM kg −1 LW); three levels of S. sesban 1198, namely, SS1 (8.2 g DM kg −1 LW), SS2 (10.9 g DM kg −1 LW), SS3 (13.3 g DM kg −1 LW) and one level of L. purpureus, LAB (10.5 g DM kg −1 LW), which was used as a control. The basal feed, chopped tef ( Eragrostis tef) straw, mineral blocks and water were offered ad libitum. The feeding period lasted for 5 months. Body weight was measured fortnightly, and monitoring of estrus was conducted twice daily for 60 days by visual observation and using vasectomized rams. Contents of crude protein (CP) were lower, and that of neutral detergent fiber (NDF), acid detergent fiber (ADF) and acid detergent lignin (ADL) were higher in L. purpureus compared to both MPTs. Soluble phenolics and condensed tannins were highest in L. pallida 14203, followed by S. sesban 1198 and least in L. purpureus. At equal levels of offer, digestible crude protein (DCP) intake was lower ( P<0.001) in sheep supplemented with L. pallida 14203 than with S. sesban 1198 or L. purpureus. L. pallida 14203 was readily consumed, whereas animals on L. purpureus and higher levels of S. sesban 1198 had variable quantities of refusals. Supplementation with L. pallida 14203 compared to S. sesban 1198 resulted in higher proportions ( P<0.01) of ewes manifesting estrus during the first estrous cycle. Manifestation of estrus was lower ( P<0.05) in ewes supplemented with the highest level of S. sesban 1198 than the two lower levels in the first and in the overall mean of the three estrous cycles monitored. Daily live weight gain was higher when the animals were supplemented with S. sesban 1198 as opposed to L. pallida 14203 ( P<0.05) or L. purpureus ( P<0.01). It is concluded that anti-nutritional factors such as condensed tannins seem to play little role in limiting intake of L. pallida 14203 at all levels of supplementation maintained in this study, whereas some anti-nutritional factors in S. sesban 1198 may limit its intake at higher levels of offer. Moreover, supplementation with S. sesban 1198 had a negative influence on the reproduction of ewes by compromising manifestation of estrus when offered at the highest level of supplementation.