The robustness of laminar wings is critical, both against instabilities that can unexpectedly trigger transition and against off-design conditions outside the cruise point. However, current inverse design methodologies not only provide suboptimal configurations but are unable to come up with robust configurations. The objective of this paper is the development and demonstration of a framework for the robust direct design of transonic natural laminar flow wings using state-of-the-art industrial tools such as computational fluid dynamics, linear stability theory, and surrogate models. The deterministic optimization problem, which serves as a baseline, searches for the optimum shape that minimizes drag applying a surrogate based optimization strategy. In that case, crossflow and Tollmien–Schlichting critical factors are fixed according to calibration data. For the robust approach, uncertainties in these critical factors as well as operational conditions are considered to account for situations that could prematurely trigger transition. The framework is able to come up with state-of-the-art natural laminar shapes for a short-haul civil aircraft configuration. The robust configurations are more balanced than the deterministic, as the transition location smoothly moves upstream as the critical factors are reduced. The pressure profiles are more resistant against instabilities, extending the current design envelope of natural laminar flow wings.