We aim to evaluate the epidemiology and outcome of gram-negative prosthetic joint infection (GN-PJI) treated with debridement, antibiotics, and implant retention (DAIR), identify factors predictive of failure, and determine the impact of ciprofloxacin use on prognosis. We performed a retrospective, multicenter, observational study of GN-PJI diagnosed from 2003 through 2010 in 16 Spanish hospitals. We define failure as persistence or reappearance of the inflammatory joint signs during follow-up leading to unplanned surgery or repeat debridement's > 30 days from the index surgery related death, or suppressive antimicrobial therapy. Parameters predicting failure were analyzed with a Cox regression model. A total of 242 patients (33% men; median age 76 years, interquartile range [IQR] 68-81) with 242 episodes of GN-PJI were studied. The implants included 150 (62%) hip, 85 (35%) knee, 5 (2%) shoulder, and 2 (1%) elbow prostheses. There were 189 (78%) acute infections. Causative microorganisms were Enterobacteriaceae in 78%, Pseudomonas spp. in 20%, and other gram-negative bacilli in 2%. Overall, 19% of isolates were ciprofloxacin resistant. DAIR was used in 174 (72%) cases, with an overall success rate of 68%, which increased to 79% after a median of 25 months’ follow-up in ciprofloxacin-susceptible GN-PJIs treated with ciprofloxacin. Ciprofloxacin treatment exhibited an independent protective effect (adjusted hazard ratio [aHR] 0.23, 95%CI, 0.13-0.40; P<0.001), whereas chronic renal impairment predicted failure (aHR, 2.56, 95%CI, 1.14-5.77; P=0.0232). Our results confirm a 79% success rate in ciprofloxacin-susceptible GN-PJI treated with debridement, ciprofloxacin, and implant retention. New therapeutic strategies are needed for ciprofloxacin-resistant PJI.This article is protected by copyright. All rights reserved.