Hydrologists do not yet possess a generally accepted catchment classification system. This article presents a review of work done so far, and discusses a general framework for a catchment classification system considering variability in relevant characteristics, increasing human impacts on catchments, and the assumption that our climate is changing. We stress that any classification system should explicitly account for uncertainty and have predictive power, rather than just being descriptive. We also propose an extension to catchment classification with the inclusion of catchment services and disservices, which would explicitly link hydrology to current and future societal issues. We discuss a framework that describes catchment climate and form and maps these on catchment function (include partition, storage and release of water, energy and matter). Climate, form, and function can be described using indices, distributions, or even conceptual models; and uncertainty needs to be preserved in individual descriptors as well as in their mapping onto each other. Descriptors of catchment function are discussed as signatures of catchment behavior, which in turn are related to catchment services and disservices. This mapping ultimately provides predictive skill by constraining the expected function at ungauged locations, even under potential nonstationarity, through knowledge of form and/or climate. Establishing this framework would provide an organizing principle, create a common language, guide modeling and measurement efforts, provide constraints on predictions in ungauged basins, and allow estimates of environmental change impacts.