This article aims to interrelate dimensions of the well-being validation instruments proposed by Watson, Clark and Tellegen (PANAS) with generalized anxiety dimensions proposed by Spitzer et al. (GAD-7) and state-trait anxiety inventories proposed by Biaggio and Natalício (IDATE), using partial least squares structural equation modeling (PLS-SEM), in the case of individual university students in southern Brazil and the city of Buenos Aires, Argentina. We conducted a behavioral study, characterized as exploratory-descriptive, by applying a questionnaire survey to collect data though face-to face interviews to a group of 460 university students from June to August 2019. A non-probabilistic sampling method for convenience was used, justified by the heterogeneous incidence of the participants. Our results support most of the proposed hypotheses. Only one hypothesis was rejected, i.e., that the Positive Affection Scale (WBS) is not related to the State Anxiety Inventory (IAE)—when a person is feeling in full activity, this situation does not affect the momentary state, characterized by tension, apprehension and by increased activity in the autonomic nervous system. In terms of the subjective well-being of students, 14.13% were found to have a low rating. 86.74% were found to have generalized anxiety; 75% had trait anxiety, and 80.22% had state anxiety. Our results indicate the need for preventive measures to minimize anxiety and help maintain necessary levels of well-being during this phase of academic development and when forging a professional career. It is expected that new studies will contribute to the advancement of such themes, particularly with university students.