Staphylococcus aureus strains can be divided into tolerant and nontolerant strains on the basis of their survival in vitro in the presence of high concentrations of methicillin (264 ,ug/ml). A strain is defined as tolerant if more than 2% of the inoculum survives under these conditions. The response of five susceptible and five tolerant S. aureus strains to treatment with methicillin was studied in an experimental thigh infection in mice. Animals were treated with one and two injections of methicillin (2.5 mg per mouse). At the end of treatment, the number of CFUs in the thigh muscles infected with the susceptible strains was found to be significantly lower than that in the thigh muscles infected with the tolerant strains.The clinical relevance of the phenomenon of tolerance has been studied repeatedly. The results reported by various authors are somewhat contradictory (2,(5)(6)(7)(8)(9) ca. 107 CFU in the left thigh muscle as stated above. Of these 18 mice, 6 were treated subcutaneously in the abdominal region with one dose of methicillin which was given 1 h after the inoculation of the bacteria. Another group of six mice received two doses of methicillin. Doses were administered at 1 and 3 h after inoculation. The remaining six mice did not receive antibiotic treatment. They were sacrificed 1 h after bacterial inoculation and served as the control group for the determination of inoculum size at the time the antibiotic treatment started. Unless stated otherwise, the doses administered were 0.15, 0.3, 0.6, 1.25, and 5 mg per mouse. Methicillin concentrations in serum. Blood specimens were obtained by orbital puncture of mice under light ether anesthesia at various intervals after the administration of methicillin. Methicillin was assayed by the aid of the agar well diffusion method, with DST agar (Oxoid Ltd., London, England) and Bacillus subtilis as an indicator organism (1).Therapeutic results. Two hours after the first and second methicillin doses (3 and 5 h after inoculation, respectively), groups of six mice (unless otherwise stated) were sacrificed by cervical dislocation. The infected thigh muscle was removed and immediately homogenized in 20 ml of ice-cold physiological saline, to which 100 RI of ,B-lactamase (15.2 U of 1-lactamase I and 2.3 U of 1-lactamase II) (Whatman Biochemicals, Ltd.) was added. The tissues were homogenized in a VirTis homogenizer for 1 min at 7,000 rpm. The homogenates were then diluted 10-fold serially in sterile saline. Each dilution (0.2 ml) was spread on nutrient agar plates, to which 0.15 U of ,B-lactamase I and 0.015 U of 1-lactamase II per ml of agar were added. After 48 h of incubation at 37°C, CFU counts were made and converted to the total number of CFUs per thigh muscle. These numbers were expressed as percentages of the mean value of the number of CFUs cultivated from muscles of untreated control mice which were sacrificed 1 h after bacterial inocu-