BackgroundDespite its high prevalence and associated levels of impairment, the latent structure of social anxiety disorder (SAD) is not well understood, with published studies reporting inconsistent results. Furthermore, it is unknown whether the latent structure of social fears in individuals with and without SAD is the same.MethodExploratory factor analysis (EFA) and confirmatory factor analysis followed by multiple indicators multiple causes (MIMIC) analysis were conducted on 13 commonly feared social situations assessed in a nationally representative sample including individuals with SAD and those with social fears but who did not meet DSM-IV criteria for SAD.ResultsAn EFA conducted in the full sample, including individuals with no social fears (88% of the sample), yielded only one factor. When the sample was restricted to those with at least one social fear, the EFA yielded three factors, in both the subsample with at least one social fear but no SAD and the subsample with SAD. The three factors represented feared situations related to public performance, close scrutiny and social interaction. The MIMIC analyses further indicated that the three-factor structure was able to explain differences in prevalence of social fears across a broad range of sociodemographic covariates.ConclusionsAmong individuals with at least one social fear and those with DSM-IV SAD the latent structure of social fears appears to be best described by three factors, although this may partially depend on how the sample is specified. These results may help reconcile the findings of different numbers of factors identified in previous studies.