Management strategy evaluation (MSE) is an analytical process used to evaluate alternative strategies for the management of renewable resource systems against explicitly stated objectives using a dynamic simulation framework. A key pre-requisite of MSE involves turning broad conceptual objectives into quantifiable and measurable operational objectives, against which the performance of management strategies can be assessed in simulations. However, given the large uncertainty typical of many renewable resource management problems and the potential diversity of stakeholder interests and needs, specification of operational management objectives for MSE often proves a challenge. In this article, a new approach to the evaluation of multidimensional outputs from MSE modeling, taking into account uncertainty regarding the reference levels of performance indicators (PIs) is proposed. The approach uses the notion of viable management strategies, as defined in recent applications of viable control to marine social-ecological systems, to examine the way in which simulated operational management objectives can be set. In this context, "viable" management strategies are defined as those which allow reference levels for candidate PIs to be met at some pre-agreed levels of tolerance, and we consider the possibility for these reference levels to be uncertain. The approach is applied to the multidimensional set of results from an MSE study conducted on recreational fishing for spangled emperor (Lethrinus nebulosus) in the Ningaloo Marine Park of Western Australia. The analysis shows how the complexities in management arrangements on recreational fishing, combining spatial management restrictions on fishing as well as conventional fisheries management regulations, which the MSE model accounted for in a multidimensional set of simulation results, can be synthesized using viability analysis. Results point to the existence of management options which provide greater "room to move", in setting reference and tolerance levels, for the range of objectives identified by stakeholders. Beyond the application to Ningaloo Reef, the approach could be transferrable to any other simulation-based outcomes of MSE for natural resource systems, both marine and land-based. © 2014 Elsevier Ltd.