The authors report the results of a randomized, prospective study to assess the effectiveness of perioperative antibiotic prophylaxis in preventing postoperative infections following clean neurosurgical operations. The study group comprised 846 patients treated between October, 1979, and June, 1984. Antibiotics, including cefazolin and gentamicin, were administered only in the immediate preoperative and intraoperative periods. Sixteen patients, none of whom developed infections, were excluded from final statistical analysis because they had inadvertently been entered into the study while failing to meet entry criteria. Fifteen wound infections (3.64%) developed in the group of 412 patients who did not receive antibiotics, whereas only four infections (0.96%) were identified among the 418 patients who received antibiotics. The difference is statistically significant (p = 0.008) and represents a 74% reduction in infection rate with antibiotics. An analysis of subgroups of surgical procedures revealed a dramatic decrease in craniotomy infections from 6.77% to 0% (p = 0.003). Of the four infections that occurred among the antibiotic-treated patients, three were in cases where foreign bodies had been implanted. No complications of antibiotic usage were identified. The rates of infection in areas of the body other than the surgical wound were no different in the antibiotic-treated and nontreated groups. All wound infections in both antibiotic-treated and nontreated patients involved similar types of Gram-positive organisms, suggesting that antibiotic prophylaxis did not produce infections with resistant or unusual organisms. This study, combined with other recently published analyses, suggests that routine perioperative antibiotic prophylaxis can significantly reduce the incidence of postoperative neurosurgical infections.