A study is reported of sepsis rates in 836 total hip replacements carried out in various physical environments including ultra-clean air. In a further sutdy the bacteriology of the air around the wound and of the wound itself was investigated during the course of 32 joint replacement operations with and without ultra-clean air. Bacteriological contamination of the air was markedly reduced by the use of ultra-clean air but the wounds themselves were no cleaner than with conventional air conditioning provided that unsterile personel were kept at least Imetri away from the sterile area. The sepsis rate was high at the outset of the study (when a physically apalling theatre was in use, the operative technique was novel and prophylactic antibiotics were not employed). Thereafter it fell and remained unchanged in spite of the introduction of ultra-clean air. We conclude that the efficacy of techniques amiced at sterilizing the air around the wound may now be unprovable in view of the low sepsis rates which can be obtained in a conventional environment. Our results do however show that unsterile personel should not be allowed to approach within Imetri of the sterile area, whatever venitlation is employed.