The Social Interaction Anxiety Scale (SIAS) and Social Phobia Scale (SPS) are 20-item companion measures of social anxiety symptoms frequently used to evaluate outcome in treatment trials. The SIAS-6, SPS-6, and Social Interaction Phobia Scale (SIPS) are promising short forms of the SIAS and SPS. The current study evaluated whether it is sound to use these short scales instead of the full-length instruments to measure outcome in social anxiety disorder (SAD) treatment studies, using data from a trial in which 255 adults with SAD were treated with traditional or imagery-enhanced group cognitive–behavioral therapy. Several deficiencies with the short forms were identified including ceiling effects, inflated variances, imprecise effect size estimates, and a loss of statistical power when testing for between-treatment differences. Using the short forms can alter the substantive findings of a treatment trial, as genuine differences in efficacy between treatments can be missed. We recommend treatment outcome be measured using the full SIAS and SPS rather than the SIPS, SIAS-6, and SPS-6. The full-length instruments provide precise estimates of treatment effects and maximize the chance of detecting between-treatment differences when they exist.