Protracted hospital-based epidemics of urinary tract infection and bacteremia due to multiply resistant gram-negative bacilli have become an increasingly common and serious problem. Failure to control such outbreaks stems partly from inability to eradicate a key reservoir, the catheterized bladder. Since eradication of bacteriuria in noncatheterized patients can be achieved with single doses of antimicrobials and correlates with urinary rather than with serum antibiotic concentrations, drugs to which an organism appears resistant by discdiffusion testing, if excreted in the urine in high concentrations, might also prove useful in eliminating catheter-associated bacteriuria. Alternatively, urinary antiseptics, for which antimicrobial sensitivity testing is not usually done, might be effective. To test this hypothesis we determined the minimum inhibitory concentrations (MICs) of 45 multiply resistant Proteus, Serratia, Klebsiella, and Pseudomonas strains isolated in 13 recent epidemics of nosocomial urinary tract infections against 10 selected antimicrobials and urinary antiseptics, and compared these MICs with expected urinary concentrations of each drug. For each genus tested, MICs for at least two antimicrobials or urinary antiseptics were well below easily achievable urinary drug concentrations. Zone size criteria often predicted which drugs had MICs below achievable urinary levels. Little difference was found between MICs determined in Mueller-Hinton broth and in urine. During an epidemic, simultaneous treatment of all patients with bacteriuria by administration of a urinary antiseptic or an antibiotic that achieves high concentrations in urine, in conjunction with brief catheter removal, might prove useful in controlling any further infection.