The Cognitive, Affective and Somatic Empathy Scales (CASES) assess three forms of empathy, each with subscales for positive and negative empathy. The present study extends this child instrument to adults and examines its factor structure and construct validity. A secondary aim is to investigate the under-researched area of positive empathy. Community samples totaling 2604 adults completed the CASES for adults, together with scales assessing construct validity. Confirmatory factor analysis supported the three-factor cognitive-affective-somatic model and a two-factor positive-negative empathy model. Findings were replicated in a second independent sample. Internal reliabilities ranged from 0.80 to 0.92. Individuals with higher psychopathy and stimulation-seeking scores were less impaired in their empathic reactions to positive relative to negative valence events, suggesting that they are relatively capable of responding emotionally to rewarding events. Somatic empathy was most strongly associated with pleasure in affective touch and with female > male gender differences in empathy. While proactive aggression was associated with reduced cognitive and affective empathy, reactive aggression was associated with increased empathy. Findings provide initial support for the utility of CASES for assessing different forms of empathy and suggest that the balance between positive and negative empathy could provide new insights into psychological traits.