Idiosyncratic drug-induced liver injury (DILI) is, like other adverse effects of drugs, underreported and underestimated in most epidemiological studies based on registries of DILI cases and reporting systems (e.g., Medwatch). The same is probably true for prospective population-based studies, although they are much more likely to mirror the true incidence of DILI. Despite these challenges, the epidemiology of DILI remains of utmost importance and is gradually coming into better focus. A recent population based study found a crude incidence of ˜19 cases per 100,000 per year. Certain agents are particularly noteworthy for their DILI risk. Amoxicillin-clavulanate continues to be the most commonly implicated agent occurring in ˜1 out of 2,300 users. Some others that standout with significantly higher risk include azathioprine and infliximab. Although statin-induced hepatotoxicity has been well documented, the risk is probably quite low. Overall, the majority of DILI in children and adults is associated with either antibiotics or anticonvulsants. Drug-induced liver injury associated with intravenously given drugs does not show any major differences from DILI due to orally administered agents. Unfortunately, our understanding of pretherapy risk assessment remains rudimentary for the most part.