The subject of drug-induced liver injury (DILI) has been evolving for decades. While various guidance and other documents have been produced to help identify and manage DILI in the clinical trial setting, as well as the clinic, there are still many aspects of the process that remain incomplete. I have selected those aspects where guidance documents either do not cover all possible scenarios or where other recommendations are open to interpretation or where controversies still exist. The following discussion includes a number of these topics, including: when is it acceptable to continue development of a drug where hepatotoxicity is observed in animal models or other preclinical assessments? Should patients with underlying liver disease be routinely included in clinical trials? Are the current clinical and biochemical stopping rules for suspected DILI appropriate for all situations? Should we still be using fold elevations based on upper limits of normal or a subject’s own baseline values to assess the level of alanine aminotransferase or other liver-associated enzyme elevations? How can we best integrate the expanding fields of toxicogenomics, pharmacogenomics, metobolomics, proteomics and other new drug and host profiling into predicting DILI? Where do we stand with respect to a DILI biomarker to replace traditional liver associated enzymes? How do we improve upon the voluntary reporting system for adverse drug reactions? What are the most useful causality assessment methodologies to diagnose DILI and is it ever possible to exclude the drug in question? And how do we best determine and manage the competing benefits and risks of an agent causing DILI?