Tissue chambers, implanted subcutaneously on both sides of the neck in eight ponies, were inoculated with Streptococcus equi subsp. zooepidemicus in order to compare the clinical efficacy of trimethoprim/sulfadiazine (TMP/SDZ) and penicillin G treatment in a purulent infection. The TMP/SDZ treatment consisted of one intravenous (i.v.) injection of 5 mg/kg TMP and 25 mg/kg SDZ and the same dose of TMP/SDZ per os (p.o.), both given 20 h after inoculation. The oral dose was then repeated every 12 h for 21 days. The penicillin treatment consisted of one i.v. injection of 20 000 IU/kg sodium penicillin G and intramuscular (i.m.) injection of 20 000 IU/kg procaine penicillin G, both given 20 h after infection. The i.m. dose was then repeated every 24 h for 21 days. Eight ponies, each with two tissue chambers, were used in a cross over design; in the first experiment the left tissue chamber (TC) was infected and in the second experiment the right. TMP/SDZ treatment resulted in a limited reduction of viable bacteria in the TC but did not eliminate the infection, resulting in abscessation in 10–42 days in all eight ponies. However, penicillin treatment eliminated the streptococci in seven of eight ponies, and only one pony suffered abscessation on day 10. This constitutes a significantly better efficacy of the penicillin treatment in this model. The most probable cause of the failure of TMP/SDZ to eliminate the streptococci is inhibition of the action of TMP/SDZ in the purulent TCF. Therefore, TMP/SDZ should not be used to treat purulent infections in secluded sites in horses.