OBJECTIVE To evaluate antimicrobial agents for treatment of models of acute and persistent leptospirosis caused by Leptospira interrogans serovar pomona. METHODS Randomized trials comparing dosages and regimens of various antimicrobial agents for treatment of acute and persistent leptospirosis. METHODS 245 Golden hamsters to model acute leptospirosis and 121 mixed-breed swine to model persistent leptospirosis. METHODS Hamsters and swine were inoculated with L interrogans serovar pomona. Antimicrobial agents were given to hamsters for 3 or 5 days after inoculation, with necropsy at 14 days after inoculation. Swine were treated for 1, 3, or 5 days beginning at 3 weeks after inoculation, and were necropsied 7 to 10 days after completion of antimicrobial agent treatment. Hamster tissue and swine tissue and urine specimens were examined by culture, fluorescent antibody testing, and histologic examination for presence of leptospires. RESULTS All untreated control hamsters became infected and manifested clinical signs and lesions of acute leptospirosis. Leptospires were not detected in hamsters treated with dihydrostreptomycin/penicillin G (25 mg/kg of body weight). Administration of ampicillin at all dosages reduced the number of hamsters infected, as confirmed at necropsy; the other agents tested required dosages greater than label recommendations to reduce the number infected. All untreated control swine became infected and shed leptospires in urine through the time of necropsy. Leptospires were not detected in kidneys or urine of swine treated with dihydrostreptomycin/penicillin G (25 mg/kg) for 1, 3, or 5 days, or in swine treated with oxytetracycline (40 mg/kg for 3 or 5 days), tylosin (44 mg/kg for 5 days), or erythromycin (25 mg/kg for 5 days). Treatment with ceftiofur and ampicillin was not effective in elimination of L interrogans serovar pomona in swine. CONCLUSIONS Dihydrostreptomycin/penicillin G is effective for treatment of acute and persistent leptospirosis. Differences between the effectiveness of antimicrobial agents in the acute and persistent model of leptospirosis emphasize the importance of using the appropriate model for treatment evaluation. Antimicrobial agents evaluated for treatment of persistent leptospirosis in swine required the use of dosages above those recommended by the manufacturer. CONCLUSIONS Use of antimicrobial agents at extra-label dosages for treatment of persistent leptospirosis may cause residue problems in food animals; however, these regimens may be useful for treatment of breeding stock or animals destined for import/export.