Neuroimaging research has strongly influenced a biologically-based conceptualization of social anxiety, which is the fear of evaluation from others. Functional neuroimaging research has shown consistently a robust association between atypical amygdala activation and social anxiety symptoms. However, there are disparities in the small structural imaging literature on the amygdala and social anxiety. The inconsistent findings may, in part, be a function of differences across studies in the methods used to obtain amygdala volumes. Freesurfer and manual tracings are two common segmentation techniques, and the use of one over the other involves different tradeoffs. The present study directly compared amygdala volumes generated based on Freesurfer’s boundaries to those generated based on manually corrected boundaries, in neurotypical adults with varying levels of social anxiety. Also, it examined whether amygdala volume predicted social anxiety symptom severity. The Liebowitz Social Anxiety Scale – Self-Report version served as a measure of social anxiety. Participants (N = 76) were selected from three larger archival projects. They had social anxiety scores ranging from 0 - 108 (M = 54.59 ± 33.34). The results suggest Freesurfer’s boundaries consistently produced larger amygdala volumes than manually corrected boundaries. However, in neurotypical individuals with and without social anxiety, manual correction did not provide added benefit over the use of Freesurfer with regard to predicting social anxiety symptoms. The present findings strongly suggest that volumetric measurement of the amygdala is not helpful for understanding variability in social anxiety symptom severity and call into question numerous aspects of existing volumetric studies of the neural correlates of social anxiety.