Migration from the primary tumor and through extra-cellular matrix (ECM), intravasation across a cellular barrier, and extravasation and recolonization in a remote site comprise the critical steps of cancer metastasis. Physiologically relevant and well-controlled models that mimic the in vivo tumor microenvironment will enable better understanding of the steps of metastasis and evaluation of potential therapy efficacy. In vivo models have physiological relevancy, yet inherently lack a high level of control. In vitro cancer models provide control, yet lack critical components of the tumor microenvironment. We propose a new technology, a microfluidic metastasis assay (uMA) that replicates essential components of the in vivo tumor microenvironment, including a 3D ECM and vasculature, while providing tight control of biochemical and biophysical parameters. The objective of the proposed work is to extend our previous work under the R21 IMAT program to further develop and evaluate our uMA. Several extensions are proposed including: (i) creating a controlled hypoxic environment, (ii) introducing realistic levels of shear stress in the vascular compartment, (iii) use of tumor spheroids to simulate EMT, and (iv) expanding the range of ECM materials currently being used (Aim 1). Another novel direction is to develop a similar assay to investigate extravasation and recolonization (Aim 2). Finally, to promote use of the uMA by other researchers and for high throughput studies, the platform is multiplexed and methods are developed for manufacturing in plastic (Aim 3). As developed, the uMA has separately addressable communicating regions for cancer cells and other tumor-associated cells seeded in a 3D collagen gel, and for endothelial cells (EC) that line a second channel to simulate the vasculature. The configuration permits migration of cancer cells from tumor spheroids within the gel toward the EC-lined channel. The EC layer mimics the in vivo vascular barrier allowing observation of cancer cell intravasation. A similar device allows cancer cells seeded in the channel to extravasate across an EC layer into ECM. Excellent optical access will permit real time observation of cancer cell migration, intravasation and extravasation. The optical access combined with image processing techniques will quantify cancer cell morphological and migratory parameters, leading to identification of novel invasion metrics that will quantify the metastatic potential of cancer cells. Finally, we will leverage the capability of the uMA for use a a functional screen for anti-metastatic drugs. These aims will establish the uMA as a useful model for quantitative research of biological mechanisms governing cancer cell metastasis. Therapies that address multiple steps of the metastatic process would clearly benefit from using the uMA as a development platform, as the system provides a well-characterized EC layer under tightly controlled microenvironmental conditions. Future development will enable the uMA to serve as a cancer cell diagnostic device and a high throughput drug development tool.