There is little research from low- and middle-income countries examining the psychometric properties of a screening tool for social anxiety disorder. The sensitivity and specificity of the Social Anxiety Screening Questionnaire against the Mini-International Neuropsychiatric Interview as a gold standard in social phobia diagnosis were investigated using analyses of receiver operating characteristics. The "best subsets" selection procedure was conducted to determine the best three to five questions. Three questions on the screening questionnaire that best discriminate between a positive and negative diagnosis of social anxiety disorder on the MINI module were identified. Answering yes to all three of these questions gives a false-positive rate of 0.44 and a false negative rate of 0.11. For this combination, the sensitivity was 0.84, and specificity was 0.67. Additional work is needed to develop a more accurate scale that could help increase the percentage of people who receive appropriate treatment of this debilitating disorder.