A retrospective study was performed in order to identify factors significantly associated with operative mortality and post-operative complications in 114 patients with liver cirrhosis who underwent resection surgery for single hepatocarcinoma nodules (diameter < or = 7 cm) from 1980 to 1992. Vascular thrombosis, extrahepatic spread of the tumor, ascites, advanced cirrhosis, severe cardio-pulmonary disease, and renal failure were exclusion criteria. Forty-one pre-operative (demographic, clinical, laboratory, and histopathological features) and two peri-operative variables (amount of blood transfused and duration of operation) were cross tabulated using chi 2 test with operative mortality and post-operative complications, which were taken as end-point variables. Stepwise logistic regression analysis was then employed in order to define factors independently associated with end-point variables. Only partial thromboplastin time abnormality was significantly (P = 0.003) and independently associated with operative mortality. The length of the operation and the period during which surgery was carried out (1980-1986 and 1987-1992) were significantly (P < 0.001) and independently associated with the occurrence of at least one post-operative complication. The results suggest that the use of more restrictive selection criteria do not make it possible to lower operative mortality, although the coagulative status should be better assessed. Minimizing surgical stress and optimizing post-operative medical care are crucial for preventing post-operative complications.