Some theorists have argued that empathy should be defined as feeling the same emotion that one thinks another person is feeling and that other-oriented sympathetic caring should be excluded from this construct. Jordan, Amir, and Bloom (2016) introduced the Empathy Index (EI) as a self-report questionnaire to measure the former circumscribed conceptualization of empathy. They reported that this scale (a) was only weakly associated with broader sympathetic caring and (b) was not robustly predictive of heightened prosociality in behavioral tasks. Using an online sample of adult community participants (N = 389), we demonstrate that the factor structure of the EI is substantively different from that proposed by Jordan and colleagues. In addition, we demonstrate that contagion for enjoyable emotions as measured by the EI is strongly correlated with sympathetic caring, interpersonal attachment, and psychological well-being, as well as substantially negatively correlated with meanness. In sharp contrast, contagion for aversive emotions, as measured by the EI, is not associated with sympathetic caring but is positively associated with emotional distress and personality disorder features. We conclude that empathic contagion should not be treated as a unidimensional construct, as enjoyable contagion and aversive contagion may be associated with markedly different nomological networks. (PsycINFO Database Record (c) 2018 APA, all rights reserved).