The ability to accurately interpret others' emotional expressions in the face, voice, and body is a crucial component of successful social functioning and has been shown to predict better outcomes in private and professional life. To date, emotion recognition ability (ERA) has mostly been measured with tests that heavily rely on static pictures of the face and on few emotions, restricting their content validity. Recently, Schlegel, Grandjean, and Scherer (Psychological Assessment, 26, 666-672, 2014) published a new test that measures ERA in a more comprehensive fashion, by (1) including a wide range of 14 positive and negative emotions and (2) using video clips with sound that simultaneously present facial, vocal, and bodily emotional cues. This article introduces the short version of the Geneva Emotion Recognition Test (the GERT-S), and presents two studies (total N = 425) that examine the internal consistency, factor structure, and convergent and discriminant validity of the test. The results show that the GERT-S is a unidimensional test with good internal consistency. Furthermore, the GERT-S was substantially positively correlated with other ERA tests, with tests of emotional understanding and emotion management, and with cognitive ability. Taken together, the present studies demonstrate the usefulness of the GERT-S as an instrument for the brief and reliable assessment of ERA. It is available, free of charge and in seven different languages, for academic research use. Given the brief test-taking time (approx. 10 min) and its possible administration via different online platforms, the GERT-S can easily be integrated by researchers into their own studies.