This article introduces the concept of “robust selection”, which proposes tree selection based on the stochastic simulation of economic values to account for the inherent uncertainty of economic weights used in tree selection for breeding programs. The proposed method uses both median ranking and ranking variability as criteria for breeding selection. Using consensus genetic and economic parameters from the New Zealand Radiata Pine Breeding Company program, we compare three selection strategies: deterministic application of economic weights from a vertically integrated bioeconomic model, an equal-weight index often used in operations, and robust selection. All strategies aim to increase value for a breeding objective that includes four traits, i.e., volume, stem sweep, branch size, and wood stiffness (measured as modulus of elasticity), based on a selection index that considers five criteria, i.e., stem diameter at breast height (1.3 m), straightness, branching score, wood density, and modulus of elasticity. Two-thirds of the selected trees were unique for each of the selection strategies. Robust selection achieved the best realised gain for three of the four selection criteria and was the middle performer in the last selection criteria. Considering the high intrinsic uncertainty of economic weights, we suggest that the relevant criterion for the selection of individuals is the maximum median ranking, subject to an acceptable level of variation in that ranking, rather than their narrow performance under a single economic scenario. This will lead to tree selections that perform well under a wide range of economic circumstances