This paper presents a decision-scaling based framework to determine whether one or more pre-selected planning alternatives for a multi-objective water-resources system are robust to a variety of non-stationary hydroclimatic conditions and modeling uncertainties. The decision-scaling methodology is advanced beyond previous applications with an efficient procedure to select realizations of climate variability and Bayesian methods to assess the effects of hydrologic uncertainty. Monte Carlo simulations are used to identify long-term planning alternatives that are robust despite the hydroclimatic uncertainties. A new metric is proposed to define robustness in this context. The framework is coupled with a host of long-term projections to understand the likelihood of potential future changes and provide useful guidance for planning. The effects of climate model down-scaling and credibility on the decision process are discussed. The approach is demonstrated in a case study for a dual-purpose surface water reservoir in Texas. The results suggest that both internal climate variability and hydrologic uncertainty can substantially alter the assessment of system robust for long-term planning purposes.