OBJECTIVE To retrospectively evaluate the outcome of treatment and identify factors prognostic for survival and locoregional and distant disease control for patients with retroperitoneal soft tissue sarcoma. METHODS The records of 104 patients with retroperitoneal soft tissue sarcoma (RSTS) managed with surgery and irradiation at Princess Margaret Hospital between 1975 and 1988 were retrospectively reviewed. Univariate log-rank analysis was used to evaluate potential prognostic factors. RESULTS Presentation was new primary disease, 74; primary recurrence, 20; metastases, 10. Pathology was liposarcoma for 42, leiomyosarcoma for 22, malignant fibrous histiocytoma for 19, and 21 with other histologies. Grade was low for 36, high for 35, and 33 were not graded. Median tumor size was 17 cm. Grossly complete surgical excision was achieved for 45 (43%), of whom 6 (6%) also had clear surgical margins. Adjuvant postoperative irradiation was administered to 36 patients to a median dose of 40 Gy/20 fractions/4 weeks and 16 received adjuvant chemotherapy. Nine patients received no adjuvant postoperative radiotherapy. Gross residual tumor was present postoperatively in 57 patients. The overall 5- and 10-year survival rates were 36% and 14%, respectively. The locoregional relapse free rate (RFR) was 28% at 5 years and 9% at 10 years, and the distant RFR was 76% at 5 years and 60% at 10 years. For the 45 patients treated with complete excision, survival was 55% and 22% at 5 and 10 years, and locoregional RFR was 50% and 18% at 5 and 10 years. Univariate analysis demonstrated that complete surgical removal was the only factor significant for improved survival, locoregional RFR, and distant RFR. Liposarcoma histology predicted for improved survival (p = 0.02), and leiomyosarcoma histology for a lower distant RFR, compared to other histologies (p = 0.003). Patients under 62 years had an improved survival (p = 0.002) and local RFR (p = 0.02), and patients presenting with recurrent disease had improved survival (p = 0.03). Sex, tumor size, or grade, or the use of adjuvant chemotherapy were not predictive for any of the endpoints tested. Those who received adjuvant irradiation following gross surgical clearance experienced a prolonged median locoregional RFR over those who did not, and this approached statistical significance for those receiving radiation doses > 35 Gy. (103 months vs. 30 months, p = 0.06). Statistical significance was reached (p = 0.02) if only the infield RFR was considered. CONCLUSIONS This study demonstrates that failure to achieve local control is the primary cause of treatment failure for patients with RSTS, and that postoperative irradiation in doses > 35 Gy after complete surgery delayed, but did not prevent local recurrence. Improvements in outcome for patients with RSTS will require alternate treatment strategies, and preoperative irradiation with an aggressive surgical attempt at complete excision is currently under investigation.