Tissue function depends on hierarchical structures that extend from single cells (~10 micron) to functional subunits (100 to 1000 microns) that in turn coordinate organ functions. Conventional cell culture disperses tissues into single cells while neglecting higher-order processes. The convergence of semiconductor-driven microtechnology with the biomedical arena now allows fabrication of microscale tissue subunits towards functionally improving in vitro models. Furthermore, as with DNA microarrays, microtechnology may revolutionize biological assays simply through the benefits of miniaturization. We have developed a miniaturized, multiwell human liver tissue model with optimized microscale architecture that maintains phenotypic functions for several weeks in vitro (as opposed to a few days in conventional culture systems). Our microscale liver model comprises of primary human hepatocytes organized in empirically optimized dimensions and subsequently surrounded by supportive murine embryonic fibroblasts. The need for such models is underscored by the high rate of pre-launch and post-market attrition of pharmaceuticals due to liver toxicity. We demonstrate utility of microscale human liver tissues in drug development through assessment of gene expression profiles, phase I and II metabolism, canalicular transport between hepatocytes, secretion of liver-specific products, and susceptibility to a panel of clinical hepatotoxins. We have spun out a company (Hepregen Corporation) and are now working with several pharmaceutical companies to define research directions which will help demonstrate the industrial utility of microscale liver tissues in both drug discovery and ADME-Tox screening. In the future, the combination of microtechnology and tissue engineering may spur development of other tissue models and their integration towards the so-called 'human-on-a-chip'.