To analyze treatment and survival in a series of resected patients with primary or recurrent retroperitoneal sarcoma (RPS) treated and prospectively followed at a single institution. Between July 1994 and December 2015, 89 patients (36 M, 53 F; mean age 60 years, range 25-79) were evaluated. For the purpose of analysis, complete resection was defined as removal of gross tumor with histologically confirmed clear resection margins. Eighty-three out of the 89 patients (93%), 46 of whom affected by primary RPS, and 37 by recurrent RPS, underwent surgical exploration. Sixty-two had a grossly and microscopically complete resection. Fifty-three out of 83 patients (64%) underwent removal of contiguous intra-abdominal organs. Preoperative mortality was nil and significant preoperative complications occurred in six cases only (7%). High-grade tumor pointed out to be a significant variable for a worse survival in all 83 patients amenable to undergo surgical resection (57% 5 years survival for low grade vs 14% for high grade; P = 0.0004). Among completely resected patients, only histologic grade clearly affected disease-free survival (72% 5 years survival for low grade vs 50% for high grade; P = 0.04), while the role of preoperative blood transfusions (67% 5 years survival for non-transfused patients vs 29% for transfused patients; P = 0.05) has to be evaluated in connection to patient complexity. Histological grade and recurrence are the most valuable prognostic predictors; in this clinical subset, an aggressive surgical approach in both primary and recurrent RPS is associated with a best long-term survival and disease-free survival.