ABSTRACT Hundreds to thousands of malignant epithelial clones with unique genetic compositions and fitness potentials are present in lethal cancers. Each clone competes against one another and also against a hostile immune microenvironment. Remarkable selective forces establish reservoirs of therapeutically resistant tumor cells and lethal metastases. Mechanistically mapping the evolution of tumor heterogeneity is the next great challenge in cancer research. Genomics approaches have been developed for digitizing heterogeneity in tumors and inferring lineage trajectories. However, gold-standard spatial information relied upon by pathologists for centuries is lost in these techniques and the inferred lineage relationships must still be proven experimentally. For almost three decades, genetically encodable protein tags have been invaluable for lineage tracing cell-fate in vivo. The discovery of green fluorescent protein (GFP) added the benefit of fluorescence microscopy. Spectrally resolvable GFP derivatives further increased the combinatorial potential of lineage tracing. Using techniques like “confetti” labelling, four clones are now commonly traced in most laboratories. Other approaches have successfully traced a hundred clones (“Brainbow”). These status quo approaches have remained unchanged for over a decade and still fall significantly short of the thousands of colors needed for quantitatively mapping tumor heterogeneity. This proposal provides a solution in the form of a massively combinatorial lineage tracing strategy in mice, termed “MousePaint”. This high-risk and high-reward IMAT project is based upon decades of genetic research from a team with an established track-record in combinatorial lineage tracing and hyperspectral imaging. MousePaint utilizes fluorescent proteins spanning the entire visible spectrum. Downstream hyperspectral imaging is used to image thousands of lineages and theoretically approach True Color imaging of >1 million colors. The objective of this proposal is to perform feasibility studies on a “MousePaint” technology that can be used to paint and visualize thousands of tumor clones during the natural history of a cancer – including initiation, growth, and metastasis. Two Aims are proposed. (Aim 1) To optimize a MousePaint strategy for combinatorial imaging in vivo and (Aim 2) To engineer and test a genetically encodable MousePaint for quantitatively visualizing the evolution of intratumor heterogeneity. Completing these aims will deliver MousePaints for imaging tumor cell heterogeneity. Hyperspectral imaging and big data analysis will benchmark MousePaint lineage tracing and color-depth. Co-registry of MousePaint with in situ transcriptomics, molecular pathology, histopathology, and scRNAseq compatibilities will be directly tested. MousePaint is expected to transform cancer research by providing a simple mouse tool for True Color imaging tumor evolution at least 100 to 1000 times the resolution of available methods.