The existence of diabetes mellitus has been postulated to have a deleterious effect on the outcome following lumbar spine surgery. We retrospectively examined the records and radiographs of 32 diabetic patients (mean age, 60 years) who underwent posterior lumbar fusions using transpedicular instrumentation and iliac crest autograft. Ten patients were insulin-dependent and 22 required oral hypoglycemic agents for at least 1 year prior to surgery. The minimum follow-up time was 2 years after surgery (mean, 2.5 years). Surgical indications included herniated lumbar disk, lumbar spinal stenosis, thoracolumbar trauma, and lumbar pseudarthrosis. Clinical results were evaluated by chart review and/or interview by using Odom's criteria. At follow-up, 75% of patients were graded as excellent or good, and 25% as fair or poor. Twenty-five of 32 patients (78%) had improvement of back pain. Twenty of 27 (74%) patients had improvement of leg pain. Eight of 15 (53%) patients had improvement in motor strength, and 6 of 11 (54%) had improvement in light-touch sensation. Insulin dependence and the presence of polyneuropathy were associated with a poorer outcome. The average time to radiographic fusion was 5 months. Twenty-nine of 32 patients (91%) developed solid fusion by strict radiographic criteria. The three patients with a pseudarthrosis had persistent back pain and a poor result. Ten of 32 (31%) of the patients experienced perioperative complications, including prolonged wound drainage (n = 5), deep wound infection (n = 1), superficial wound infection (n = 1), atrial fibrillation (n = 1), ruptured cerebral aneurysm (n = 1), and ulnar nerve neuropathy (n = 1). We conclude that posterolateral lumbar spinal fusion with internal fixation in diabetic patients yields clinical results comparable to those of nondiabetic patients, with similar risks of perioperative complications.