Clinically, the stiffness and shape of a solid tumor are important diagnostic markers of malignancy, and physical palpation remains a major method in cancer detection. Rapid cell proliferation and extracellular matrix (ECM) deposition within confined tissue spaces lead to tumor stiffening and the buildup of solid stress, which exists in two primary forms: compressive and tensile. While the effects of tensile stress, including cell-generated traction forces, have been extensively studied, our understanding of compressive stress is limited. Although compressive stress has been implicated in tumor progression, mechanistic understanding is lacking, largely due to the lack of experimental platforms capable of simultaneously measuring tumor mechanics and capturing single-cell resolution imaging in physiologically relevant 3D environments. In this thesis, I developed a microfluidic rheometer that applies static or cyclic compression to tumor spheroids embedded in 3D collagen matrices while enabling real-time imaging of tumor spheroid deformation and single-cell dynamics. A key innovation is the integration of a polyacrylamide membrane force sensor, which allows precise mechanical measurements in a biologically relevant and optically accessible environment. Applying this tool to breast cancer models, we uncovered two key findings. First, direct stress-strain measurements enabled the first quantitative characterization of the Young’s modulus and viscoelastic properties of tumor spheroids, establishing the Microrheometer as a microscale equivalent to conventional rheometers. Second, by combining mechanical loading with real-time imaging, we correlated tumor mechanics with single-cell dynamics, leading to the discovery of tissue fluidization under compression and highlighting the potential link between invasiveness and tumor mechanics. The Microrheometer provides a powerful, high-throughput platform for characterizing tumor biomechanics and invasion in 3D environments. Its versatility also enables future studies on single-cell and nuclear mechanics, making it a valuable tool for mechanobiology research and potentially clinical diagnostics. In addition to the development of the Microrheometer, this dissertation presents microfluidic platforms developed to study tumor responses to other biophysical (e.g., ECM glycation) and biochemical (e.g., EGF gradient) cues within the tumor microenvironment, revealing new insights into how environmental factors drive tumor progression.