We hypothesized that health of the donor nerve and corresponding muscle, as assessed by electromyography (EMG), could predict the outcome of nerve transfer surgery. A retrospective review was performed to investigate outcomes of nerve transfers for elbow flexion and shoulder abduction. Motor strength was graded preoperatively and after a minimum 1-year follow-up. Preoperative EMG results were classified as functionally normal or affected based on motor unit recruitment pattern and correlated with follow-up motor strength and range of motion. Forty nerve transfers were identified: 27 were performed for elbow flexion and 13 for shoulder abduction. Overall, the 29 transfers in the normal EMG cohort showed significantly greater postoperative improvement in motor strength (Medical Research Council grade 0.2-4.1) than the 11 transfers in the affected EMG cohort (grade 0.0-3.0). In the shoulder cohort, normal donor nerves resulted in greater strength (grade 4.0 vs. 2.4) and active motion (83° vs. 25°) compared with affected donor nerves. Double fascicular transfers with 2 normal donor nerves demonstrated improved strength compared with double nerve transfers when 1 donor nerve was affected (grade 4.5 vs. 3.2). Our findings demonstrate that a simple EMG classification that describes the quality of donor nerves can predict outcome as measured by postoperative motor strength and range of motion. Preoperative EMG evaluation should be considered a valuable supplementary component of the donor nerve selection process when planning brachial plexus reconstruction. Prognostic II.