Daptomycin is an attractive option for treating prosthetic joint infection, but the 6-mg/kg of body weight/day dose was linked to clinical failure and emergence of resistance. Using a methicillin-resistant Staphylococcus aureus (MRSA) knee prosthesis infection in rabbits, we studied the efficacies of high-dose daptomycin (22 mg/kg given intravenously [i.v.] once daily [o.d.]; equivalent to 8 mg/kg/day in humans) or vancomycin (60 mg/kg given intramuscularly [i.m.] twice daily [b.i.d.]), both either alone or with adjunctive rifampin (10 mg/kg i.m. b.i.d.). After partial knee replacement with a silicone implant, 10(7) MRSA CFU was injected into the knees. Treatment started 7 days postinoculation and lasted 7 days. Positive cultures were screened for the emergence of mutant strains, defined as having 3-fold-increased MICs. Although in vivo mean log(10) CFU/g of daptomycin-treated (4.23 ± 1.44; n = 12) or vancomycin-treated (4.63 ± 1.08; n = 12) crushed bone was significantly lower than that of controls (5.93 ± 1.15; n = 9) (P < 0.01), neither treatment sterilized bone (2/12 and 0/12 rabbits with sterile bone, respectively). Daptomycin mutant strains were found in 6/12, 3/12, and 2/9 daptomycin-treated, vancomycin-treated, and control rabbits, respectively; no resistant strains emerged (MIC was always <1 mg/liter). Adjunctive rifampin with daptomycin (1.47 ± 0.04 CFU/g of bone [detection threshold]; 11/11 sterile bones) or vancomycin (1.5 ± 0.12 CFU/g of bone; 6/8 sterile bones) was significantly more effective than monotherapy (P < 0.01) and prevented the emergence of daptomycin mutant strains. In this MRSA joint prosthesis infection model, combining rifampin with daptomycin was highly effective. Daptomycin mutant strains were isolated in vivo even without treatment, but adjunctive rifampin prevented this phenomenon, previously found after monotherapy in humans.