The current monograph asserts that nonverbal behaviors can serve as useful overt indices of experienced arousal. Various conceptualizations of arousal and factors eliciting arousal in communication situations are examined. It is proposed that, from a communication standpoint, indices of arousal and arousal change are best classified along two dimensions of arousal intensity and arousal valence. Diverse bodies of literature on nonverbal concamitants of arousal are reviewed. As a first test of whether separate profiles exist for negatively, as opposed to positively or neutrally, valenced arousal, data are analyzed from a mock-interview experiment in which, following a baseline interview, participants (N=52) were subjected by an interviewer to either increased or decreased involvement. The involvement changes were expected to elevate arousal intensity for all participants and induce positively or negatively valenced arousal in the two respective conditions. Polynomial regressions and z-test comparisons of correlations revealed that 16 composites, comprising a total of 47 nonverbal variables, showed significant associations with changes in arousal.