Liver toxicity has been one of the major causes for drug attrition. It has caused discontinuation of preclinical and clinical studies, and withdrawals and black box warnings of marketed drugs. Drug-induced hepatotoxicity may lead to organ failure and, ultimately, patient death. Understandably, liver toxicity is one of the most important issues in the drug development cycle. In silico models based on chemical structure provide a cost effective approach for hepatotoxicity, and it can help prioritize lab tests, preclinical and clinical studies. There is no doubt as to the need for predictive methods in every stage of the drug discovery pipeline. However there is a serious shortage of consistent data for the development of reliable models and developing genuinely predictive models for hepatotoxicity remains a challenge.