The experiment had two objectives: 1) to determine the protein requirements of two strains of growing-finishing pigs based on growth performance, carcass characteristics, tissue accretion rates, and organ weights and 2) to evaluate whether protein requirements can be determined from changes in plasma urea concentration. Forty-six Gene Pool (GP) and 46 Hampshire (H) gilts with an initial BW of 28.5 kg were used. Pigs were allotted to two trials of a randomized complete block experiment with a 2 x 6 factorial arrangement of treatments. Five pigs from each strain were randomly selected and slaughtered at the beginning of each of the two trials. The remaining 72 pigs were individually penned and allotted to one of six dietary treatments (10, 13, 16, 19, 22, or 25% CP). Pigs remained on the experiment until the mean weight of a treatment group within each strain reached 115 kg (16 wk for GP and 14 wk for H), at which time all pigs of that strain were slaughtered. The only strain x protein level interactions that were detected were for carcass protein and water accretion rates. Gene Pool pigs grew less rapidly and utilized feed less efficiently than H pigs (P < .001). Average daily gain (quadratic, P < .05) and ADG/ADFI (quadratic, P < .05) were increased as protein level increased until a plateau was reached. Backfat depths were decreased (linear, P < .001) and longissimus muscle areas were increased (linear, P < .001) as protein level increased. Protein accretion rate was lower (P < .01) and fat accretion was higher (P < .01) in GP pigs than in H pigs. Protein accretion increased (quadratic, P < .001) and fat accretion decreased (linear, P < .001) with increasing dietary protein level. Examination of the response of plasma urea concentration over time suggested that GP pigs required 13% CP from 30 to 80 kg and 10% CP thereafter, whereas H pigs required 19% CP from 30 to 45 kg, 16% CP from 45 to 100 kg, and 13% CP thereafter.