Two experiments were carried out to evaluate the effect on broiler perfoman- ce and the economics of increasing by 5 o/o the metabolizable energy (ME) in diets containing cassava meal (CM: 0, 20 and 30 o/o), by adding either vegetable oil (Experiment 1) or tallow (Experiment 2). One day old mixed sex Hubbard broiler chicks were randomly assigned to each of six treatment groups composed of four replicates per treatment. Each pen constituted an experimental unit with 36 and 37 chickens for Experiments 1 and 2, respectively. A commercial (Exp. 1) and sorghum- based (Exp. 1 and 2) diets were used as controls. During the starter period diets containing 20 and 30 o/o CM, without any incremental of ME, produced higher live weight and better feed conversion than the control. In Experiment 2 diets containing 30 o/o CM yielded lower performance than diets with 20 o/o CM. The inclusion of 20 o/o CM yielded higher live weight at 7 (Experiment 2) or 8 (Experiments 1 and 2) weeks than the other diets, whereas the 30 o/o CM diets showed similar results to those of the control diets. The addition of vegetable oil or tallow improvement of feed conversion, especially during the starter period. Diets containing 20 o/o CM yielded the highest returns.