In this study I develop theoretically the role of emotions in identity theory by examining individuals' emotional reactions to identity nonverification (in a positive and a negative direction) and identity verification, which occurs once versus repeatedly, and which is perpetrated by a familiar other compared with an unfamiliar other. Predictions from identity control theory (ICT) are used to guide the analysis. An experiment simulates a work situation and invokes the worker role identity. Workers either receive feedback that is expected, given their worker identity standard (identity verification); feedback that is more positive than they would expect (a lack of identity verification in a positive direction); or feedback that is more negative than they would expect (a lack of identity verification in a negative direction). The workers' emotional reactions' to each situation are investigated. Contrary to ICT, identity nonverification in a positive direction results in positive (not negative) emotions; the persistence of verifying and nonverifying feedback decreases (rather than increases) the affective response to the feedback; and feedback from a familiar other does not significantly influence actors' emotional reactions. The findings raise some questions about current thinking in ICT and suggest important extensions for emotions in identity theory. (Contains 7 tables and 21 footnotes.)