Wood frogs, like other amphibian species worldwide, are experiencing population declines due to multiple stressors. In the northeastern United States, wood frog declines are thought to result from a reduction in successful metamorphosis in part due to climate change, disease (specifically ranavirus) and contaminant exposure. The presence of multiple stressors can increase uncertainty in characterizing the main effects of each stressor, as well as understanding the degree to which their effects interact (additively or synergistically) to impact populations. This uncertainty adds inherent challenges to selecting appropriate management actions for conserving populations. Finding solutions that are robust to these uncertainties can improve management amid absent or equivocal knowledge. We used a Bayesian Belief Network (BBN), a quantitative tool that allowed us to evaluate how potential management actions might mitigate the effects of increasingly frequent and severe droughts, ranavirus exposure, and methylmercury on wood frog populations in the northeastern US. In our system, successful wood frog recruitment was largely driven by hydroperiod regardless of other stressors. Our modelling indicated that increased hydroperiod lowered the probability of complete metamorphosis failure from 0.6 to 0.37, suggesting that under the conditions tested in the model, pond hydrology, is more important for successful recruitment than either methylmercury or ranavirus exposure. As more information becomes available on stressor interactions, model scenarios could be re-run and management options re-evaluated.