"Chills" (frisson manifested as goose bumps or shivers) have been used in an increasing number of studies as indicators of emotions in response to music (e.g., Craig, 2005; Guhn, Hamm, & Zentner, 2007; McCrae, 2007; Panksepp, 1995; Sloboda, 1991). In this study we present evidence that chills can be induced through aural, visual, tactile, and gustatory stimulation. Additionally, chills were elicited by mere mental self-stimulation--even without any external stimulus. Subjective ratings and physiological responses (i.e., skin conductance response, heart rate, breathing rate) of 36 participants reporting chills were recorded in response to stimuli of the named sensory domains. Chills in response to all external stimuli showed similar physiological correlates. However, they differed in the subjective affective perception: more chills occurred in response to "negative" valent and arousing sounds and pictures, while more chills were reported for "positive" valent music independent of arousal. These findings suggest that the chill phenomenon could be a valuable indicator of strong emotions. However, the measurement of chills should be combined with other methods of emotion measurement, since chills can be a response to psychological events of different affective value. (Contains 4 figures and 4 tables.)