The ability to mentalize (i.e., to form representations of mental states and processes of oneself and others) is often impaired in people with schizophrenia spectrum disorders. Emotional awareness (EA) represents one aspect of affective mentalizing and can be assessed with the Levels of Emotional Awareness Scale (LEAS), but findings regarding individuals with schizophrenia spectrum disorders are inconsistent. The present study aimed at examining the usability and convergent validity of the LEAS in a sample of N = 130 stabilized outpatients with schizophrenia or schizoaffective disorders. An adequacy rating was added to the conventional LEAS rating to account for distortions of content due to, for example, delusional thinking. Scores of the patient group were compared with those of a matched healthy control sample. Correlation with symptom clusters, a self-report measure of EA, a measure of synthetic metacognition (MAS-A-G), and an expert rating capturing EA from the psychodynamic perspective of psychic structure (OPD-LSIA) were examined. Regarding self-related emotional awareness, patients did not score lower than controls neither in terms of conventional LEAS nor in terms of adequacy. Regarding other-related emotional awareness, however, patients showed a reduced level of adequacy compared to controls whereas no such difference was found for conventional LEAS scores. Higher conventional LEAS scores were associated with fewer negative symptoms, and higher structural integration of self-perceptions measured by the OPD-LSIA. Higher adequacy of responses correlated with fewer symptoms of disorganization as well as excitement, higher scores of self-reflection on the MAS-A-G as well as self- and object-perception and internal and external communication as measured by the subscales of the OPD-LSIA. Findings suggest that the LEAS might not be sensitive enough to detect differences between mildly symptomatic patients with schizophrenia or schizoaffective disorders and healthy controls. However, LEAS ratings are still suitable to track intraindividual changes in EA over time. Observing the adequacy of patients’ responses when using the LEAS may be a promising way to increase diagnostical utility and to identify patterns of formal and content-related alterations of mentalizing in this patient group. Methodological indications for future studies are discussed.