<p><em></em>The dietary preference and voluntary intake of fruits from <em>Acacia cochliacantha</em>, <em>Caesalpinia cacalaco</em>, <em>Vachellia pennatula</em>, <em>Chloroleucon mangense</em>, <em>Senna atomaria</em> and <em>Guazuma ulmifolia</em> [containing 8.5-13.6% crude protein (CP) and 25.9-72.9% neutral detergent fibre (NDF) on dry matter (DM) basis], and the partial replacement of a commercial diet with 0, 15, 30 and 45% G. <em>ulmifolia</em> fruit were studied in rabbits. Rabbits fed a standard diet (17.9% CP and 28.4% NDF) were used to study their preference (during 20-min periods) for 6 ground fruits (Experiment 1: 12 rabbits, from 45 to 56 d of age and 1180±114 g live weight [LW]), 6 pelleted fruits (Experiment 2: same rabbits from 57 to 68 d of age), and the voluntary intake of the 5 most preferred pelleted fruits (Experiment 3: 18 rabbits from 70 to 81 d of age and 2200±200 g LW). Then, we examined the inclusion of <em>G. ulmifolia</em> fruit up to 45% in the diet (Experiment 4: 12 rabbits/diet from 25 to 66 d of age, and 419±80 g initial LW). When ground fruits were offered, rabbits tended to prefer <em>C. mangense</em> (0.83±0.12 g DM; P&lt;0.05), and when offered as pellets they preferred <em>G. ulmifolia</em> (4.50±0.47 g DM; P&lt;0.05), as well as in Experiment 3 (36.9±3.11 g DM/d; P&lt;0.05). Substitution of a commercial diet with 15, 30 and 45% <em>G. ulmifolia</em> fruits in Experiment 4 resulted in higher DM intake when fruits were included in the diet, across all experimental periods (25-38, 39-66 and 25-66 d; P&lt;0.001). Feeding up to 30% of fruits improved growth rate from 25 to 38 d of age (P&lt;0.001) and produced similar growth rates to the control treatment from 25 to 66 d of age. Feed efficiency was impaired above 15% substitution level during the 25-38 d period (P&gt;0.05) and in all groups fed <em>G. ulmifolia</em> in the 25-66 d period (P&lt;0.001). <em>Guazuma ulmifolia</em> appears to have the greatest potential as a supplementary feed for rabbits in the pelleted form. Its inclusion up to 15% in the diet might render a similar performance to that of rabbits fed commercial diets.</p>