From January 1991 to June 1997, patients undergoing primary elective monolateral or bilateral total knee replacement (TKR) were consecutively enrolled in a prospective, open clinical study on the efficacy and safety of regional prophylaxis with teicoplanin (TEC). Those scheduled for monolateral TKR (115 patients) received 400 mg of TEC in 100 ml of saline as a 5-min infusion into a foot vein of the leg to be operated on immediately after the tourniquet was inflated to 400 mm Hg (ca. 50 kPa). For patients undergoing bilateral surgery (45 patients), regional administration of TEC was also repeated for the second knee operation. Follow-up ranged from a minimum of 2 years to 8 years. None of the patients experienced local or systemic adverse effects following regional administration of TEC. In the immediate postoperative and 2-year follow-up periods, only one superficial infection of the primary site attributable to intraoperative contamination (prophylaxis failure) out of the 205 prostheses implanted was observed. Deep infections involving the prosthesis did not occur. Infectious complications at distant sites were observed in nine cases (urinary tract infection due to Escherichia coli in eight cases, and Salmonella enteritidis gastroenteritis in one case) in the immediate postoperative period; they all were rapidly cured after antibiotic treatment. A delayed prosthetic infection, related to hematogenous spread of the etiological agent and therefore not considered a prophylactic failure, was observed in a patient who had undergone TKR 5 years before. Regional administration of TEC in monolateral and bilateral TKR appears to be a safe and valuable prophylactic technique.