Drug-induced liver injury (DILI) is a major cause of attrition during drug development and post-market withdrawal. Early identification of DILI risk during lead optimization will result in selection of clinical candidates with more favourable safety attributes. However, current in vitro models of hepatotoxicity are limited to assessing acute cytotoxicity only, while the onset of DILI is often observed following several weeks of treatment. Recently, human liver microtissues or spheroids consisting of primary hepatocytes and Kupffer cells have been shown to preserve liver specific functions over several weeks in culture, allowing longer-term repeat dose assessments, not currently possible with 2D-hepatocyte culture systems. We have recently published the predictivity of 3D human liver microtissues to discriminate between DILI and non-DILI causing drugs, utilizing ATP depletion as an indicator of organoid viability. We assessed >100 marketed drugs with known DILI potential up to 14 days exposure. The time required for formation of spheroids by human hepatocytes is subject to donor-to-donor variability, in addition to the usual differences in drug metabolism and transport – hence primary human hepatocytes are considered a poor tool for early toxicity screening. We have developed a hepatic cell line-(HepG2, clone C3A)-based spheroid model as a high-throughput in-house, early hepatotoxicity screen with broadly similar predictivity to primary hepatocyte spheroids. Finally, these models form part of a continuous, probabilistic statistical model allowing risk visualisation from human-based in vitro models, incorporating uncertainty and variation.