Introduction Despite decades of research on how people with social anxiety evaluate themselves and others, it remains unclear whether people who evaluate themselves negatively also evaluate others negatively. Findings from other-evaluation research are equivocal, perhaps attributable to methodology differences and inconsistent operationalization. Social-cognitive and cognitive-behavioural models suggest that negative self-evaluations may cause participants to subsequently evaluate a visibly anxious person negatively. We tested this hypothesis experimentally, using a video-recorded social interaction and novel false-feedback manipulation. Methods: 169 unselected participants completed baseline questionnaires and a 10-min impromptu conversation task with a confederate. We randomly assigned participants to receive positive, ambiguous, or negative false-feedback about their performance. Next, they evaluated their own performance and watched a recorded conversation between an anxious and confident speaker. Finally, they evaluated the anxious person’s performance. Results: Our manipulation was effective; participants in the negative-feedback condition rated themselves more negatively. However, no differences emerged between conditions on most cognitive and emotional outcomes. Discussion: Evaluating oneself negatively, on its own, may not lead people to evaluate a visibly anxious person in a recorded social interaction negatively in a single-session experiment within an unselected sample. Future studies should examine this relationship with a clinical sample across time and contexts.