To perform a psychometric analysis of the Brazilian version of the Brief Social Phobia Scale (BSPS).

Hundred and seventy-eight university students of both genders aged on average 21.2 years and identified as Social Anxiety Disorder (SAD) cases and non-cases was studied, with the structured clinical interview for DSM-IV being used as a parameter. The different instruments were applied in an individual manner in the presence of a rater and of an observer.

The BSPS showed adequate internal consistency (0.48–0.88) and concurrent and divergent validity with the Beck Anxiety Inventory (BAI) (0.21–0.62), Social Phobia Inventory (0.24–0.82) and Self Statements During Public Speaking Scale (SSPS) (0.23–0.31). Discriminative validity revealed a sensitivity of 0.88–0.90 and a specificity of 0.81(0.83 for cut-off notes of 18/19. Factorial analysis demonstrated the presence of six factors that jointly explained 71.79% of data variance. Construct validity indicated some limits of the scale regarding the diagnosis of SAD. Inter-rater reliability was strong (0.86–1.00, p < 0.001).

The BSPS is adequate for use with university students, although further studies in different cultures, samples and contexts are still necessary.