OBJECTIVE: To distinguish sensitivity analysis and uncertainty analysis, characterize their differential roles in health economic decision making, and to provide practical examples of their use and presentation in health economic analysis. METHOD: The role of one-way sensitivity analysis is to quantify the impact of varying a single parameter on the output of a model. However, this obscures an important distinction between parameter uncertainty and variability. Sensitivity analysis quantifies parameter variability in terms of the percentage change in a model output for a given percentage change in a model input. Sensitivity is therefore an objective property of the model. Uncertainty analysis, on the other hand, propagates a decision maker’s subjective parameter uncertainty through a model to estimate the conditional uncertainty of the model output. Accordingly, the functional role of sensitivity analysis is to help a decision maker to understand and validate the internal model structure in order to gain trust in the model itself; whereas the functional role of uncertainty analysis is to assess the potential impact of a decision maker’s subjective parameter uncertainty on confidence in a particular model-based decision. These distinctive roles are both critical in health economic analysis and decision making. We provide examples of sensitivity analysis versus uncertainty analysis, show how to report the results of sensitivity and uncertainty analyses, and discuss the implications of this distinction for conducting one-way and probabilistic analyses. CONCLUSION: Confidence in model-based decision making requires 1) confidence in the model itself, and 2) confidence in the model output given one’s subjective parameter uncertainty. Sensitivity analysis and uncertainty analysis, respectively, serve these differential roles.