Our knowledge of the cellular changes that lead to liver cell carcinoma in humans is limited by proper and necessary ethical restriction on clinical research. We know rather more about risk factors, the most important of which is cirrhosis, it seems that both the causative agent and the time of duration of the cirrhotic process are relevent to the magnitude of this risk. According to present knowledge, alpha1-antitrypsin deficiency, alcoholism, naturally occurring carcinogens, drugs, and the hepatitis B virus seem to carry the greatest risk of cancer developing in a cirrhotic patient. Cirrhosis, however, is not an essential prerequisite, and some or possibly all of these agents can also induce cancer without cirrhosis. Bile duct carcinoma commonly follows infestation with liver flukes. Cirrhosis is usually absent but duct epithelial hyperplasia is present prior to the development of cancer. Many cellular changes have been observed in patients and among populations considered to be at risk from liver cancer. Of these, liver cell dysplasia is the most striking and studies of its prevalence, natural history, and association with cirrhosis suggest that it is a precancerous change.