Glucocorticoids (CORT) are frequently used in conservation as indicators of stress in wildlife. However, the context-dependent nature of CORT means that CORT levels may not always accurately reflect condition. For example, there is growing evidence that populations can evolve or acclimate to human-induced environmental change (i.e. pollutants) by expressing higher tolerance levels. Mechanisms that allow for populations to achieve higher pollutant tolerance may affect CORT and thereby its viability as an indicator of condition. To date, no studies have considered this context. We ask 1) do populations that differ in tolerance to pollutants differentially express baseline and stress-induced CORT and 2) is CORT a viable indicator of condition across populations that differ in tolerance to pollutants? Towards this goal, we identified 3-NaCl tolerant and 3-NaCl susceptible populations of wood frogs. We reared larvae from these populations to metamorphosis in either an environmentally relevant concentration of NaCl or a control. At metamorphosis we used a non-invasive waterborne assay to measure baseline and stress-induced CORT release rate and measured fitness metrics. Regardless of the rearing environment, tolerant populations had lower baseline CORT release rates and responded to an acute stressor by mounting a positive CORT response. In contrast, susceptible populations had higher baseline CORT release rates and decreased CORT in response to an acute stressor. Thus, while CORT levels differ across populations that vary in relative pollutant tolerance, the CORT-fitness relationship across populations was consistent with the CORT-Fitness hypothesis. Collectively, this underscores the utility of CORT as a tool for evaluating wildlife condition that is robust to shifts in population traits.