Non-alcoholic fatty liver disease (NAFLD) is becoming increasingly prevalent in developed countries and there are currently no approved therapeutic strategies to ameliorate its effects. While a number of pharmaceutical companies have approached the problem, current in vivo studies used to evaluate new therapeutics are limited by low throughput, expense, ethical concerns, and the poor ability of many animal models to mimic the disease state. While humanized in vitro assays can address many of these concerns, the choice of model and assay format--2D vs 3D--requires a balance of throughput, cost, and translative capacity. While 2D models are higher throughput and less expensive, standard culture approaches can limit the sensitivity of these assays, while co-culture approaches to more fully recapitulate the many stages of the disease are more challenging to implement. On the other hand, 3D cell culture models can feature multiple cell subtypes and may offer improved sensitivity and superior recapitulation of disease state. This presentation will describe both 2D and 3D cell culture approaches to modeling the NASH disease phenotype and will discuss approaches to quantifying specific end-points associated with disease progression including inflammation, collagen deposition, and cell death.