THE VALUE OF antibiotic prophylaxis for clean neurosurgical procedures without the implantation of a foreign body has been conclusively demonstrated. Attempts to confirm its efficacy for cerebrospinal fluid shunt operations have produced confusing and inconclusive results. The objective of this study was to combine the results of high-quality controlled trials of antibiotic prophylaxis for cerebrospinal fluid shunt operations and to determine if there is evidence for the efficacy of this policy. Randomized clinical trials identified from presentations at national meetings and in the published literature were subjected to a metanalysis. The pooled data suggest a statistically significant effect favoring antibiotic prophylaxis (approximately a 50% reduction in infection risk when antibiotic prophylaxis is used). The effect is strongly related to the baseline infection rate when prophylaxis is not used and disappears when the baseline infection rate is at or below about 5%.