Treatment of end stage renal disease (ESRD) patients by renal transplant is severely limited by shortage of donor organs, while dialysis is expensive, inconvenient, and confers significant morbidity and mortality. There are nearly 400,000 people in the US who rely on thrice-weekly, in-center hemodialysis, and collectively, this population consumes over $20 billion annually in Medicare-paid healthcare. The prevalence of ESRD is increasing at 5% per year, and the vast majority of patients are unlikely to ever receive a transplant. We recently embarked on the development of an implantable bioartificial kidney that combines a hemofilter constructed from silicon nanopore membranes (SNM) with a bioreactor of human renal tubule cells to mimic nephronal function. In the final envisioned implementation, blood will be filtered in the hemofilter under circulatory system pressure to remove uremic toxins, salts, small solutes, and water. The resulting ultra filtrate will then be processed by the bioreactor to selectively transport most of the salts, and water back into the blood, thereby maintaining volume homeostasis and electrolyte balance. Initial pilot studies supported by a NIH/NIBIB-sponsored Quantum Grant (1R01EB008049) allowed our team to establish fundamental concept feasibility including the development of high-performance SNM filters, anti- fouling thin-film polymer coatings, human renal tubule cell isolation and expansion techniques, and short-term implantable hemofiltration in rodents and pigs as well as wearable cell therapy in sheep. We also identified a number of critical roadblocks to successful development of implantable bioartificial kidney. Among them, a key challenge is long-term blood compatibility of the hemofilter with respect to thrombosis and membrane fouling. The proposed R01 project will attempt to better understand the blood-device interactions spanning across anatomic, histologic, and molecular length scales and their influence on hemofilter biocompatibility. More specifically, we will conduct transport characterization experiments, computational fluid dynamics (CFD) simulations, in vitro radiographic flow mapping, and in vivo animal experiments to evaluate the impact of membrane physicochemical properties on mass transfer characteristics and determine the role of fluid flow anomalies in device thrombosis. Beyond the immediate application to an implantable bioartificial kidney, this work will establish a new generalized testing methodology for implantable devices that are functionally dependent on features at both large (mm-cm) and small (nm-microns) length scales.