There are no biological marks to indicate if denervated muscle atrophy after nerve injury is irreversible. Clinically in obstetric brachial plexus palsy (OBPP), atrophy of denervated intrinsic musculature of the hand is much faster to irreversible than that of denervated muscles of the arm. 64 pup rats whose C5C6 had been divided and C7C8T1 avulsed, were divided equally into the reconstruction and control groups. The former had subgroups R1, R5, R10, R15 where the ulnar and musculocutaneous nerves were reconstructed one, five, ten and 15 weeks respectively after injury and efficacy was evaluated 12 weeks later. The latter had C1, C5, C10, C15 subgroups where denervated muscles of the two nerves were assessed one, five, ten and 15 weeks after injury. Results of average cross-sectional area of the muscle fiber for intrinsic musculature of the forepaw showed that the R5, R10, R15 subgroups were not statistically superior to the C5, C10, C15 ones, respectively, though R1 was; those for biceps indicated, however, that the R1, R5, R10 subgroups were better than the C1, C5, C10 ones, respectively, though R15 was not. In the reconstruction subgroups regenerative nerve fibers in each nerve were no less than 53 percent of those on the control side, while number of motor end plates was statistically less in subgroups with irreversible muscle atrophy. We conclude that rat model of OBPP is suitable for simulating clinical appearance of atrophy of denervated intrinsic musculature of the hand being faster than that of denervated muscles of the arm.