Drug induced liver injury (DILI) is a major health problem in the United States and accounts for the majority of clinical holds and postmarketing use restrictions by the FDA. The majority of adverse liver reactions are idiosyncratic and their underlying mechanisms are still not well understood. Better predictive models for iDILI would enable the preclinical elimination of drug candidates with hepatotoxic liabilities. We have previously developed a model in which primary hepatocytes (rat, human, dog, or monkey) are seeded onto ECM-coated domains of optimized dimensions and subsequently co-cultivated with murine embryonic fibroblasts (HepatoPacTM). Hepatocytes in HepatoPac retain their in vivo-like morphology, express a complete complement of liver-specific genes, metabolize compounds using active Phase I/II drug metabolism enzymes, secrete diverse liver-specific products, and display functional bile canaliculi for several weeks in vitro. Here, we supplement the HepatoPac co-cultures with primary Kupffer macrophages for use in evaluating cytokine- drug interactions. Drug- induced stresses may interact with cytokine signalling to cause hepatotoxciity. Previous work using an in vivo mouse model demonstrated that LPS-induced inflammatory stress rendered trovafloxacin (TVX), but not its non-toxic analog, levofloxacin (LVX), hepatotoxic by a TNF-a-mediated mechanism. Using TVX as a prototype compound, we evaluated the ability of our HepatoPac- Kupffer cell co-culture model to detect drug/ cytokine hepatic toxicity synergies.