We used 525 bullfrog tadpoles, distributed into 15 boxes with one tadpole per liter. The water was daily renewal (200%). The feeding regimes were constituted of five arrangements using seven commercial rations with levels of 22, 28, 32, 36, 40, 45 and 55% of crude protein (CP), supplied every 15 days. We used a random blocks design, subdivided plots with three replications. The subplots were constituted of five biometries: at the beginning of the experiment, at 15, 30, 45, and 60 days. The tadpoles submitted to feeding regime FR1, with 22, 32, 36, and 40% of CP, presented the highest consumption and the same performance as the others. Tadpoles that received FR5, with 40, 45, 45 and 50% of CP, presented greater weight, however, they showed smaller weight gain, greater consumption, conversion and mortality, indicating that this regime is expensive and not adequate for management. Feeding regimes FR2, with 28, 32, 36 and 40% of CP, and FR3, with 32, 36, 40 and 45% of CP, were more adjusted to this kind of management. Animals in all treatments showed higher performance that animals fed traditionally.