There is growing evidence that subcellular compartmentation allows signaling pathways to perform multiple functions while maintaining input-output specificity. For example, cyclic AMP signaling in the heart regulates contractility, metabolism, apoptosis and hypertrophy in a receptor-specific manner. Quantitative models and experimental approaches are needed to understand precisely how subcellular compartmentation of signaling is achieved, and how such compartmentation affects specificity and other signaling properties. I will discuss two examples of how we have combined computational models and live-cell imaging to elucidate mechanisms and consequences of signaling compartmentation. Compartmentation of cyclic AMP signaling is widely recognized to be regulated by localized degradation by phosphodiesterases. However computational models by our group and others have predicted that phosphodiesterases are not sufficient: physical barriers are also required to quantitatively explain cyclic AMP compartmentation. We have developed a new technique to directly assess the role of physical barriers in cyclic AMP compartmentation. Fluorescently labeled (and phosphodiesterase resistant) cyclic AMP was microinjected into live cardiac myocytes. By both FRAP and kymograph analysis of diffusing fluorescent cyclic AMP, we measured substantially reduced diffusion than expected from its molecular weight alone. Disrupting microtubule or especially actin cytoskeleton accelerated cyclic AMP diffusion, indicating that cytoskeletal structures help restrict cyclic AMP diffusion. Formation of protein complexes provides another major mechanism for signaling compartmentation. A-kinase anchoring proteins (AKAPs) localize protein kinase A and other signaling proteins to discrete locations, and AKAPs have been shown to regulate the specificity of both cardiac hypertrophy and contractile signaling. While most mathematical models assume free diffusion of signaling proteins, we propose a new “scaffold state switching” model of signaling within a protein complex. Using the tethering of protein kinase C (PKC) to AKAP7α as an example, we predicted and then experimentally validated acceleration and amplification of PKC signaling by AKAP7α. Further, the model predicted that AKAP7α insulates PKC from ATP- and substrate-competitive inhibitors, but not activation-competitive inhibitors. These model predictions were validated experimentally as well. The generality of the predicted amplification, acceleration and insulation by AKAPs was examined using model sensitivity analysis. Overall, these two examples illustrate how mathematical modeling can be integrated with live-cell imaging to determine the mechanisms and functional consequences of subcellular signaling compartmentation.