Objectives To determine whether the incidence of bacteremia after transrectal ultrasound-guided prostate biopsy (TRUSGPB) significantly diminishes with the setting up of a new preventive protocol. This protocol was set up after detecting an augmented incidence of bacteremia after TRUSGPB with a high prevalence of antibiotic-resistant microorganisms. Methods Retrospective descriptive and prospective intervention study performed at a University Hospital. Participants: Patients undergoing TRUSGPB under the old preventive protocol (January 2006-February 2007), that is, amoxicillin-clavulanate 500 mg tid the day before, the day of the procedure, and 1 day after the procedure, and after setting up a new protocol (March 2007-April 2008), that is, 2 g cefoxitin 1 hour before the procedure and ciprofloxacin 750 mg p.o. bid the day before, the day of the procedure, and 3 days after the procedure; dipstick urinalysis was performed before the procedure, and patients with positive results were not biopsied. Results Incidence of bacteremia with old and new protocols: 9 of 204 procedures (4.4%) vs 2 of 207 (0.9%), ( P = .03). Four isolates (44.4%) under the old protocol produced extended-spectrum beta-lactamase (ESBL). With the new protocol, 2 (0.9%) cases of non-ESBL Escherichia coli bacteremia were observed. Sixty-five (23.8%) cases were not biopsied because of positive result of dipstick urinalysis, lack of antibiotic prophylaxis adherence, or altered coagulation parameters. Conclusions Antibiotic prophylaxis for TRUSGPB should take into account local resistance patterns. Cefoxitin could be used as prophylaxis in centers with high prevalence of ESBL enterobacteriaceae. Before TRUSGPB, excluding patients with positive results of dipstick urinalysis is an advisable practice.