The clinical findings in 130 conseucutive cases of spinal cord compression by metastatic extradural tumors were analyzed. These 130 patients were combined with a previous survey of 105 patients to compare the effectiveness of radiation therapy (RT) alone with that of surgical decompression followed by RT. Ambulation after treatment was considered a successful outcome. The most common primary tumors producing spinal cord compression were (in order) breast, lung, prostate, and kidney. In 68% of these tumors the thoracic region was involved. Pain was the primary symptom of 96% of the patients, while motor or sensory deficits (or both) were found in 82% of them. Therapy consisted of surgery and RT in 65 patients and RT alone in 170 patients. There were no differences in outcome between those treated by surgery combined with RT and those managed by RT alone. Patients with radiosensitive tumors and those ambulatory at the onset of treatment benefited whether treated by surgery or by RT. Seventy-five percent of living patients who improved from treatment remained ambulatory at 6 months, and approximately 50% of living patients were ambulatory at 1 year. We conclude that RT without decompressive laminectomy is as effective as decompressive laminectomy in treating epidural spinal cord compression from systemic cancer.