Two operating teams (25 persons) were followed for two months with fingerprint samples taken preoperatively; before and after 'in-use' surgical handwashing; and immediately postoperatively, with and without surgical gloves. The mean time for handwashing for the cardiothoracic team (CT) was 2 min and for the orthopaedic team (OT) was 3.5 min. A closer observation of 10 persons revealed a great individual variation in washing techniques, in spite of standard guidelines. The CT team performed eight, and the OT team nine sterile operations with an average duration of 3 h and 20 min and 2 h and 40 min, respectively. Surgical handwashing resulted in fingertip sterility in 111/118 (94.1%) cases; in 61/66 (92.4%) samples from the surgeons and in 50/52 (96.2%) samples from the assistants. Postoperative fingerprinting with gloves on showed sterile conditions in 85/91 (93.4%) samples; 57/59 (96.6%) from the surgeons and 28/32 (87.5%) from the assistants. Immediately after removal of the gloves, 43/67 (64.2%) of fingerprint samples from the surgeons and 13/48 (27.1%) from the assistants were still sterile. Coagulase-negative staphylococci (CNS) and Bacillus species predominated in fingerprint samples. Of the 105 CNS strains tested, 11.4% were methicillin resistant. Only five strains of Staphylococcus aureus were isolated; in 4/5 cases from the OT. This study illustrates that in spite of standard guidelines, there is great individual variation in surgical handwashing. However, in most instances, the bacteria are eradicated from the fingertips. Even after surgery for 2-3 h, there may still be a residual effect of the hand disinfecting agent in half of the cases.