Twenty-five febrile patients with a history of intravenous drug use who were receiving either vancomycin (15 patients) or teicoplanin (10 patients) as part of a multicenter, double-blind, randomized, clinical efficacy trial were enrolled, upon receipt of their first dose of antibiotic, into a study to evaluate the effect of 1 g of vancomycin and high-dose teicoplanin (30 mg/kg of body weight) on histamine release and the occurrence of "red man syndrome" (RMS). In addition, 10 healthy volunteer subjects (HVS) were randomized to receive either 1 g of vancomycin intravenously or a saline infusion in a double-blind, crossover design study. Patients and HVS were observed for the presence of erythema, flushing, pruritus, and hypotension during and for up to 1 h postinfusion by a blinded investigator. Histamine concentrations in plasma were measured at baseline and during and after drug infusion. No significant differences were noted in baseline temperature between patients (vancomycin recipients, 102.3 degrees F [39.1 degrees C]; teicoplanin recipients, 102.4 degrees F [39.1 degrees C]) or incidence of bacteremia (7 of 15 vancomycin recipients; 5 of 10 teicoplanin recipients). There were no significant differences in peak vancomycin concentrations in the sera of patients (40.8 micrograms/ml) and HVS (49.9 micrograms/ml). There were no reactions consistent with RMS in any patient who received teicoplanin (0 of 10); there was a significant difference in the occurrence of RMS in patients in comparison with that in HVS (0 of 15 patients, 9 of 10 HVS; P less than 0.001) who received vancomycin. The predominant reaction was erythema and pruritus. Histamine concentrations in plasma and the area under the histamine plasma concentration-time curve were highly variable within groups and were not statistically different between patients and HVS. The incidence of RMS secondary to vancomycin or teicoplanin in our patient population appears to be low and consistent with clinical observations. Similar to previous investigations, RMS secondary to vancomycin in HVS was high (90%). However, we found no relationship between the histamine concentration in plasma or the area under the plasma histamine concentration-time curve and the severity of RMS in HVS. The reason for the discrepancy of RMS in patients versus that in HVS in unknown, but it may be related to a blunted effect of glycopeptides to produce the reaction in the presence of infection or it may be specific to our patient population.