There is no universally agreed definition of well-being as a subjective experience, but Huppert and So (2013) adopted and systematically applied the definition of well-being as positive mental health-the opposite of the common mental disorders described in standard mental health classifications (e.g., Diagnostic and Statistical Manual of Mental Disorders). We extended their theoretical approach to include multi-item scales, using 2 waves of nationally representative U.S. adult samples to develop, test, and validate our multidimensional measure of well-being (WB-Pro). This resulted in a good-fitting a priori (48-item, 15-factor) model that was invariant over time, education, gender, and age; showed good reliability (coefficient αs .81-.93), test-retest correlation (.73-.85; M = .80), and convergent/discriminant validity based on a multitrait-multimethod analysis, and relations with demographic variables, selected psychological measures, and other multidimensional and purportedly unidimensional well-being measures. Further, we found that items from 2 widely used, purportedly unidimensional well-being measures loaded on different WB-Pro factors consistent with a priori predictions based on the WB-Pro factor structure, thereby calling into question their claimed unidimensionality and theoretical rationale. Because some applications require a short global measure, we used a machine-learning algorithm to construct 2 global well-being short versions (five- and 15-item forms) and tested these formative measures in relation to the full-form and validity criteria (to download short and long versions see https://ippe.acu.edu.au/research/research-instruments/wb-pro). The WB-Pro appears to be one of the most comprehensive measures of subjective well-being, based on a sound conceptual model and empirical support, with broad applicability for research and practice, as well as providing a framework for evaluating the breadth of other well-being measures. (PsycINFO Database Record (c) 2020 APA, all rights reserved).