Thoracic epidural analgesia (TEA) enhances recovery after bowel surgery. Early postoperative prolonged-release oral formulation of oxycodone or oxycodone/naloxone is potentially useful as a second analgesic step to reduce the duration of TEA. We hypothesized that oxycodone would decrease the duration of TEA and combined with naloxone preserve gastrointestinal function. Ninety patients undergoing open cystectomy and urinary diversion were enrolled in this randomized double-blind, three-arm, parallel-group, placebo-controlled single-center trial between September 2015 and February 2017. Exclusion criteria were known allergy to oxycodone/naloxone, pulmonary diseases, hepatopathy, and analgesics nonnaïve patients. From postoperative day 3, patients received batches with oxycodone, oxycodone/naloxone, or placebo every 12 hours (n = 30 in each arm). Reduction of the epidural drug infusion rate was attempted with the goal to maintain a pain intensity <3 at rest and <5 (numeric rating score) at mobilization during 6 hours. Primary endpoint was duration of TEA and secondary endpoint return of gastrointestinal function. The median duration of TEA did not differ between patients treated with oxycodone/naloxone (6.7 [range 3.1-10.3] days), oxycodone (7.0 [3.0-9.1]), or placebo (6.4 [3.1-8.4]); P = 0.88. Time to the first defecation was prolonged in the oxycodone group compared to the placebo group (difference 22.48 hours ±8.95; P = 0.037). In the oxycodone group, we found 8/30 patients with ileus (27%) compared to 2/28 (7%) in the oxycodone/naloxone group and to 2/30 (7%) in the placebo group; (P = 0.031). Oxycodone, with or without naloxone, did not reduce the duration of TEA. Oxycodone alone led to a delayed return of bowel function, whereas the combination was not different from placebo.