Watersheds are spatially explicit landscape units that contain a range of interacting physical, ecological and social attributes. They are social–ecological systems that provide a range of ecosystem services valued by the society. Their ability to provide these services depends, in part, on the degree to which they are impaired by human‐related activity. An array of indicators is used by natural resource managers, both private and government, to assess watersheds and their sub‐components. Often these assessments are performed in comparison with a reference condition. However, assessments can be hampered because natural settings of many systems differ from those sites used to characterize reference conditions. Additionally, given the ubiquity of human‐related alterations across landscapes (e.g. atmospheric deposition of anthropogenically derived nitrogen), truly unaltered conditions for most, if not all, watersheds cannot be described. Definitions of ‘integrity’ have been developed for river ecosystems, but mainly at the reach or site scale and usually for particular species, such as fish or macroinvertebrates. These scales are inappropriate for defining integrity at the watershed scale. In addition, current assessments of endpoints do not indicate the source of impairment. Our definition of watershed ‘integrity’ is the capacity of a watershed to support and maintain the full range of ecological processes and functions essential to the sustainability of biodiversity and of the watershed resources and services provided to society. To operationalize this definition as an assessment tool, we identify key functions of unimpaired watersheds. This approach can then be used to model and map watershed integrity by incorporating risk factors (human‐related alterations or stressors) that have been explicitly shown to interfere with and degrade key functions in watersheds. An advantage of this approach is that the index can be readily deconstructed to identify factors influencing index scores, thereby directly supporting the strategic adaptive management of individual components that contribute to watershed integrity. Moreover, the approach can be iteratively applied and improved as new data and information become available.