As part of an investigation carried out in 50 hospitals we particularly checked the sterilisers available for proper functioning and correct operation. The work group examined 461 sterilisation programmes in 338 sterilisation units (176 autoclaves, 153 hot-air apparatus and 9 ethylene oxide units) during the daily routine operation. The group carried out biological and physical inspections and carefully recorded the faults and deficiencies detected during the sterilisation procedure. It was found that only in 22% of the hospitals was a proper functioning and correct operation of all the sterilising units in use ensured. Technical faults and errors in operating the equipment and especially in the loading in 23% of the programmes prevented sterility. Almost 50% of the hot-air sterilisers and about a quarter of the autoclaves fail to produce a consistent satisfactory effect. The incidence of errors which occurred with ethylene oxide sterilisation cannot be generalised, since only a limited number of units were checked. A breakdown by functional areas and wards indicates that serious errors occur not only in the centralised large units but also in little-used small autoclaves in peripheral areas. A comparison with the literature shows that our results do not differ to any appreciable extent from those obtained by other workers, over the last twenty years. Here systematic inspections by external control authorities could improve the situation within a few years. Regular inspections to be carried out by the hospitals themselves using modern test methods are necessary, and need internal record-keeping. In spite of this, we cannot at present do without external inspections. We pin our hopes mainly in regular and thorough expert training of the staff and on much closer cooperation between hospital and hygienist.