ABSTRACT Mechanistic models of biological processes can help explain observed phenomena and predict response to a perturbation. A mathematical model is typically constructed using expert knowledge and informal reasoning to generate a mechanistic explanation for a given observation. Although this approach works well for simple systems with abundant data and well-established principles, quantitative biology is often faced with a dearth of both data and knowledge about a process, thus making it challenging to identify and validate all possible mechanistic hypothesis underlying a system behavior. To overcome these limitations, we introduce a Bayesian multimodel inference (Bayes-MMI) methodology, which quantifies how mechanistic hypotheses can explain a given experimental datasets, and concurrently, how each dataset informs a given model hypothesis, thus enabling hypothesis space exploration in the context of available data. We demonstrate this approach to probe standing questions about heterogeneity, lineage plasticity, and cell-cell interactions in tumor growth mechanisms of small cell lung cancer (SCLC). We integrate three datasets that each formulated different explanations for tumor growth mechanisms in SCLC, apply Bayes-MMI and find that the data supports model predictions for tumor evolution promoted by high lineage plasticity, rather than through expanding rare stem-like populations. In addition, the models predict that in the presence of SCLC-N or SCLC-A2 cells, the transition from SCLC-A to SCLC-Y through an intermediate is decelerated. Together, these predictions provide a testable hypothesis for observed juxtaposed results in SCLC growth and a mechanistic interpretation for tumor recalcitrance. AUTHOR SUMMARY To make a mathematical model, an investigator needs to know and incorporate biological relationships present in the system of interest. However, if we don’t know the exact relationships, how can we build a model? Building a single model may include spurious relationships or exclude important ones, so model selection enables us to build multiple, incorporating various combinations of biological features and the relationships between them. Each biological feature represents a distinct hypothesis, which can be investigated via model fitting to experimental data. We aim to improve upon the information theoretic framework of model selection by incorporating Bayesian elements. We apply our approach to small cell lung cancer (SCLC), using multiple datasets, to address hypotheses about cell-cell interactions, phenotypic transitions, and tumor makeup across experimental model systems. Incorporating Bayesian inference, we can add into model selection an assessment of whether these hypotheses are likely or unlikely, or even whether the data enables assessment of a hypothesis at all. Our analysis finds that SCLC is likely highly plastic, with cells able to transition phenotypic identities easily. These predictions could help explain why SCLC is such a difficult disease to treat, and provide the basis for further experiments.