The role of the combination of perioperative antibiotics, laminar air-flow operating rooms, and body-exhaust systems in reducing the incidence of deep-wound sepsis after total hip replacement remains controversial. We evaluated the incidence of deep-wound infection as associated with the magnitude of the operation in 575 patients (659 procedures) who had all three methods of prophylaxis. The same surgeon performed all of the operations. The procedures were divided into five groups, according to the complexity of the operation and the need for major bone-grafting. Only one patient (who had previously had an infection) had early sepsis. After the 376 standard primary operations, the over-all incidence of sepsis (early and late) was 0.38 per cent. No infection occurred after fifty-four complex primary operations without femoral-head autografting. After fifty-nine complex primary operations with femoral-head autografting, three patients had a late deep infection. No infection occurred within two years after the 170 revision operations. There was no late infection after the 104 revisions without major bone-grafting, but late sepsis developed in two patients after the sixty-six revisions with major bone-grafting. Of the total of six cases of late infection in five patients, five had an identifiable source of infection that was extrinsic to the wound. The incidence of early postoperative sepsis after total hip replacement, even complex primary total hip replacement and revision, was extremely low when using the three contemporary methods of prophylaxis for infection.(ABSTRACT TRUNCATED AT 250 WORDS)