From January to May 1982, 37 children undergoing CSF shunt procedures in two different countries were enrolled in a randomized, double-blind, controlled study to receive vancomycin hydrochloride (15 mg/kg/dose) or placebo (saline) one hour before surgery and again six hours later. Twenty patients received vancomycin, and 17 received placebo. In the 35 cases that could be evaluated, shunt-associated infections developed in three (17%) of 18 patients who received vancomycin and in four (23%) of the 17 placebo recipients. All infections were caused by Staphylococcus species susceptible to vancomycin. A histaminelike rash developed in seven (35%) of 20 patients during vancomycin infusion. It recurred with readministration in one patient and was accompanied by hypotension in another patient. The reactions were not related to too rapid infusion of vancomycin. Because of the adverse reactions to vancomycin, the study was discontinued.