An experiment was carried out to evaluate the effect of including the optimum level (20 %) of cassava flour in diets for feed chicken in combination with three sources of protein, on the biological yields and their economic results. The experimental varieties were made up of 4 diets (treatments) which varied in terms of the source protein: diet 1, 0% of cassava flour, and cottonseed meal, diet 2 with 20% cassava flour and three sources of protein (soybean meal, fish flour and cottonseed meal, diet 3 with 20% cassava flour and two sources of protein (soybean meal and fish flour), and diet 4 with 20% cassava flour and one source of protein (soybean meal). Results for the complete period showed statiscally similar performance for live weight gain, feed intake, and feed conversion between diets 1 and 2. Diet 3 although showing no statistical differences with the other treatments, resulted in biological terms less live weight gain. Diet 4 always showed the lowest liveweight gains and feed intake. Economic evaluation showed maximum net benefits for diet 2.