Potential assessment has served various objectives in the development of renewable energies. However, the prospective nature of this type of assessment sometimes makes it difficult to evaluate and compare estimation results based on different data and modeling. To facilitate this comparison, uncertainty estimates need to be systematically provided. Since potential assessment sometimes relies on numerous parameters, this may first require determining the most important inputs to focus on. In this paper, we propose a sensitivity analysis methodology based on Sobol indices so as to identify the main inputs from a nonlinear assessment model. We illustrate the proposed methodology through analyzing sensitivity in an onshore wind and ground solar photovoltaic (PV) potential assessment covering two French regions. As a result, we show that, when estimating the potential of these renewable energy sources, parameters defining surface availability are more prevalent than those related to technology. (C) 2016 Elsevier Ltd. All rights reserved.