Different theoretical models exist to conceptualize well-being, with most of them proposing correlated facets of well-being that combine to define a flourishing life. Standard confirmatory factor analytic (CFA) approaches do not account for two theoretically expected sources of construct-relevant multidimensionality, namely the hierarchical nature of the constructs and the imperfect nature of items that may produce cross-loadings on nontarget factors. These limitations are addressed by bifactor CFA, exploratory structural equation modeling (ESEM) and bifactor ESEM. This presentation will suggest the value of these psychometric developments for enhanced measurement of well-being, and, importantly, for a better understanding of well-being on a theoretically level. Illustrations will be based on data from different culturally diverse South African samples that completed the Mental Health Continuum Short Form (Keyes, 2006, 2009). This scale operationalizes Keyes’s Mental Health Continuum model of positive mental health which specifies emotional, social, and psychological well-being as constituents. Repeatedly, the sophisticated models resulted in improved model fit with the overall scale but not the subscales displaying adequate reliability. The psychological well-being items mainly represented overall mental health in an indigenous African sample, but not in other samples. The implications for well-being theory and measurement across cultural contexts will be discussed.