The aim of this study was to biologically assess three protein sources used in commercial dog food: soy meal (TS), meat meal (HC), and chicken meal (HP) using four biological tests: Protein Efficiency Ratio (PER), True Digestibility (DV), Net Protein Utilization (NPU) and Biological Value (VB). 40 albino laboratory rats of 23 days of age were used. Five groups were formed according to the protein source: groups TS, HC, HP, control group (sodium caseinate) and a non protein-group. The results were: PER: TS = 1.86, HC = 1.39, HP = 1.69, control = 2.11; DV: TS = 98.76, HC = 96.95, HP = 98.13, control = 99.19; NPU: TS = 58.19, HC = 35.52, HP = 60.19, control = 96.46; VB: TS = 0.59, HC = 0.37, HP = 0.61, control = 0.97. PER results in the HC group were significantly lower than other groups (p<0.05). There was not statistical differences on PER between groups HP and TS. All protein sources had a high true digestibility. The HC was the protein source of less quality because the lowest PER, NPU and VB values. It is concluded that to asses the quality of protein nutrients is not enough to establish the protein content and digestibility, but to conduct biological tests.