Nineteen patients were evaluated 6--17 months after medial epicondylectomy (22 extremities) for ulnar nerve compression syndrome at the elbow. Medial epicondylectomy removes the prominence against which the ulnar nerve can be traumatically compressed and no operative handling of the nerve is necessary. Pre- and postoperative nerve conduction velocities were combined with subjective and objective evaluations of all patients. Statically significant correlations were found between the result and the postoperative nerve conduction velocity and the result and the per cent change of nerve conduction velocity. Ten extremities were classified as good results (48%), 4 as fair results (17%), and 8 as poor results (35%). Four limbs in the poor category and one limb in the fair category had electrical evidence of proximal compression neuropathy which compromised the end result. Exclusion of the patients with evidence of proximal neuropathy would yield 76% good or fair results. Measurement of ulnar nerve lengths in fresh cadavers demonstrated that ulnar nerve conduction velocity studies should be performed at 70 degrees of elbow flexion if accurate pre- and postoperative velocities are to be obtained.