This prospective study was performed to determine whether gentamicin can be prescribed routinely to patients with spinal cord injury undergoing urologic procedures, or whether antibiotic therapy must be selected on the basis of recent urine microbiologic test results. Between January 2004 and June 2005, procedures were performed on 38 patients, all of whom were prescribed antibiotics on the basis of a microbiology report. Sixteen patients who underwent urologic surgery during 2003 and received gentamicin empirically served as a control group. The patients' clinical course was monitored for postprocedure sepsis. Only 12 patients received gentamicin as the sole antibiotic; 10 patients required an additional antimicrobial for urine samples that grew more than I organism and contained bacteria resistant to gentamicin; 26 patients needed antibiotics other than gentamicin for gentamicin-resistant uropathogens. Three patients with organisms sensitive to gentamicin as well as another antibiotic received an agent considered less nephrotoxic than gentamicin. After the procedure, sepsis occurred in only I patient, a man with chronic lymphocytic leukemia and small cell carcinoma of the urinary bladder. Three control group patients developed a fever in excess of 39 degrees C. One of these patients did not require a change of antibiotic, another patient recovered after 3 changes of antibiotic, and the third patient recovered from septicemia after receiving ventilatory support. Antibiotics should be prescribed on the basis of recent urine microbiologic test results, and empiric therapy with gentamicin should be avoided in patients with spinal cord injury who are scheduled to undergo urologic procedures.