Butterbur extracts (Petasites hybridus) are recommended for the prevention of migraine, but pharmacovigilance reports may be suggestive of rare hepatobiliary toxicity. To evaluate its hepatotoxic potential, a series of in vivo and in vitro studies were carried out. Essentially, there were no signs of hepatocellular toxicity at estimated therapeutic C(max) levels of 60 ng/ml. Nonetheless, in a 28-day toxicity study at approximately 200-fold of therapeutic doses, induced liver transaminases and bilirubin elevations were observed. In a subsequent 6-month chronic toxicity study, the initial hepatobiliary effects were reproduced, but at the end of the study, liver function recovered and returned to normal as evidenced by clinical chemistry measurements. To identify possible mechanisms of hepatotoxicity, we investigated liver function in vitro at > 170-fold of therapeutic C(max) levels, including cytotoxicity (lactate dehydrogenase, MTT, and ATP), transaminase activities (alanine aminotransferase and aspartate aminotransferase), albumin synthesis, urea and testosterone metabolism to assay for cytochrome P450 monooxygenase activity. Only with extracts rich in petasin (37% petasin) and at high and well above therapeutic doses, liver toxicity was observed. A toxicogenomic approach applied to hepatocyte cultures enabled hypothesis generation and was highly suggestive for extracts high in petasin content to impair bile acid transport and lipid and protein metabolism. Importantly, neither chronic rat in vivo nor rat in vitro studies predicted reliably hepatotoxicity, therefore reemphasizing the utility of human-based in vitro investigations for the development of safe medicinal products. Finally, toxicogenomics enabled the characterization of a novel butterbur extract with no signals for hepatotoxicity.