Tissue microenvironments affect the functional states of cancer cells, but determining these influences in vivo has remained a challenge. We present a quantitative high-resolution imaging assay of single cancer cell morphology in zebrafish xenografts to probe functional adaptation to variable cell-extrinsic cues and molecular interventions. Using cell morphology as a surrogate readout for functional cell state, we focus on Ewing Sarcoma, a pediatric cancer driven by the oncogene EWSR1-FLI1, as a prototypical cancer whose adaptation to microenvironments likely drives disease outcome through non-genomic mechanisms. Computer vision analysis reveals systematic shifts in the distribution of 3D morphotypes as a function of cell type and seeding site, as well as tissue-specific cellular organizations that recapitulate those observed in human tumors. Reduced EWSR1-FLI1 causes a shift to more protrusive cells and decreased tissue-specificity of the morphotype distribution. Overall, this work establishes a framework for the statistically robust study of cancer cell plasticity in diverse tissue microenvironments.