What is a nonverbal expression of emotion? Both the notion of expression and the notion of emotion are contentious in the literature. Everyone knows the clear cases – smiles, frowns, screams, chuckles, slumps, and so on – but the category as a whole is not well defined. Writers from different theoretical backgrounds have criticized the implicit assumptions inherent in this phrase (Ekman, 1971; Hinde, 1985; Parkinson, 2005; Zajonc, 1985). Are the referenced behaviors in fact expressing something, and is this something an emotion? Not all scientific accounts are consistent with the implication that certain nonverbal behaviors express an emotion.However, for simplicity of reference we will continue to use the phrase “nonverbal expression.” But we do so in an inverted-commas sense only, namely, to refer to those nonverbal behaviors that are commonly taken to express emotions. We acknowledge that the category is vague, and we remain agnostic on whether what is expressed is truly an emotion, or, indeed, whether “express” is what such behaviors do.We are similarly agnostic on the definition of emotion, and we do not use that word here in any technical sense. Instead, our focus in this chapter is on short-term emotion episodes, which we take to be multi-componential events of limited duration commonly taken to be an emotion. Components include but are not limited to appraisals, physiological changes, subjective experiences, nonverbal expressions, and instrumental behaviors.We now turn to summarizing how the basic emotion, appraisal, and psychological constructionist research programs account for the production and perception of nonverbal expressions. (See chapters in this volume by Shiota, Ellsworth, and Barrett, respectively, for more general discussions of each research program and for fuller sets of references.) Although each program is commonly called a theory, they are instead broad research programs: each includes a family of loosely related (indeed sometimes conflicting) theories and assumptions, an interpretation of the history of the field, various background assumptions about human nature, prescribed methods and data analytic procedures, and conclusions drawn from previous research. Furthermore, each program continues to develop. We present a prototypical version of each program, emphasizing differences among the three research programs. That said, the three research programs also share important assumptions, methods, and conclusions, although the emphasis may vary. For example, when we describe one program’s account of evolutionary origins, the reader should not infer that the other two programs reject evolution by natural selection or assume special creation. Similarly, the fact that one program emphasizes context does not mean that contextual effects are incompatible with the other programs. We present each program’s assertions as if they were established facts, but in fact they are hypotheses. In the conclusion to our chapter, we elaborate on compatibilities and convergences, but we begin by contrasting the three programs.