Abstract Current in vitro platforms are poor predictors of the in vivo safety, efficacy and pharmacokinetics of therapeutics, owing to a significant difference in the test conditions compared to physiological conditions. Therefore, drug toxicity testing is routinely performed using animal models. However, animal testing is expensive and time consuming. In addition, ethical concerns about the use of animals are increasingly calling for reduction/replacement of animal tests. To overcome these challenges, physiologically relevant organ-on-chip assays have been developed. These assays mimic the dynamic interactions encountered during drug delivery and recapitulates physiological flow rates, vascular architecture and the 3D nature of tissue (liver, lung, kidney, etc.), thereby providing improved quantitative and predictive capabilities to guide the development of drugs via accurate toxicity analysis. However, one of the critical components lacking from current organ-on-chip assays is the real-time analysis of drug concentration at specified locations within the assay to determine drug toxicity at defined tissue sites. To address this need, we propose to integrate our microfluidics-based, organ-on-chip systems with on-chip mass spectrometry analysis to measure drug concentrations across a vascularized liver construct. The Phase I effort will focus on integration the microfluidic device with a novel mass spectrometry (MS) assay. This method enables online temporal and spatial chemical characterization of chemical constituents within microfluidic devices by MS for the first time. The ChemSitu approach enables the means to continuously sample and chemically characterize small volumes of liquid directly from a microfluidic device at any point along the construct in near real-time and without negatively altering the state of the microfluidic system. A multi-disciplinary team of scientists and engineers with expertise in microfluidics-based cell assays and instrumentation development has been assembled for successful completion of this project. By providing an accurate, quantitative and predictive model of and quantitation of physiological interactions, the developed platform promises to establish a new paradigm for in vitro assessment of the physiological response to therapeutics.