A total of 11 cases of red man syndrome collected among 650 children who had received vancomycin in our hospital between 1986 and 1988 (estimated prevalence 1.6%) were retrospectively analyzed. These 11 children were compared with 11 age-matched children who received vancomycin in whom red man syndrome did not develop. Of the patients with red man syndrome, 73%, and of the patients with no reaction, 45.4% received vancomycin for penicillin-resistant Staphylococcus epidermidis-positive cultures, or because of history of penicillin allergy. No difference was observed in the dose per kilogram given to both groups (12.9 +/- 3.5 mg/kg per dose in those with red man syndrome vs 12.3 +/- 6.9 mg/kg per dose in control children. The duration (mean +/- standard deviation) of vancomycin infusion was 45.9 +/- 16.7 minutes (range 10 to 90 minutes) in patients with red man syndrome and 54.5 +/- 7.6 minutes (range 45 to 65 minutes) in the control group (P = .07). In the 5 children with red man syndrome rechallenged with vancomycin, slower infusion rates prevented or reduced the syndrome, which emphasized the fact that the rate of administration is the important determinant of red man syndrome in susceptible cases. Clinically, the syndrome developed at the end of the infusion in most patients, but appeared as early as 15 minutes after initiation of the infusion. It was mostly manifested as a flushed, erythematous rash on the face, neck, and around the ears. Less frequently, the rash was distributed all over the body. Pruritus was usually localized to the upper trunk but was also generalized (2 of 11 children).(ABSTRACT TRUNCATED AT 250 WORDS)