Much research has suggested that those who stutter are likely to be anxious. However, to date, little research on this topic has addressed the role of expectancies of harm in anxiety, which is a central construct of anxiety in modern clinical psychology. There are good reasons to believe that the anxiety of those who stutter is related to expectancies of social harm. Therefore, in the present study, 34 stuttering and 34 control participants completed the Fear of Negative Evaluation (FNE) Scale and the Endler Multidimensional Anxiety Scales-Trait (EMAS-T). The FNE data showed a significant difference between the stuttering and control participants, with a large effect size. Results suggested that, as a group, a clinical population of people who stutter has anxiety that is restricted to the social domain. For the EMAS-T, significant differences between groups were obtained for the two subtests that refer specifically to people and social interactions in which social evaluation might occur (Social Evaluation and New/Strange Situations) but not for the subtests that contained no specific reference to people and social interactions (Physical Danger and Daily Routines). These results were taken to suggest that those who stutter differ from control subjects in their expectation of negative social evaluation, and that the effect sizes are clinically significant. The findings also suggest that the FNE and the EMAS-T are appropriate psychological tests of anxiety to use with stuttering clients in clinical settings. The clinical and research implications of these findings are discussed, in terms of whether social anxiety mediates stuttering or is a simple by-product of stuttering. Possible laboratory explorations of this issue are suggested, and potential Cognitive Behavior Therapy packages for stuttering clients who might need them are discussed. Educational objectives: The reader will be able to: (1) explain why expectancy of social threat or harm may be associated with stuttering; (2) name and describe two psychological tests that are suitable for assessment of the social threat or harm that may be associated with stuttering; and (3) explain how findings for the EMAS-T test in the present results suggest that expectancy of social threat or harm, but not other kinds of negative expectancy, are associated with stuttering. (Contains 1 table.)