Escitalopram has demonstrated efficacy for the acute treatment of social anxiety disorder (SAD) in two placebo-controlled trials and for long-term treatment in a relapse-prevention study. Social anxiety disorder is a heterogeneous disorder. This study questions whether this new selective serotonin reuptake inhibitor is effective across different subgroups of patients. Data from two randomised, placebo-controlled, 12-week escitalopram SAD trials were pooled. General linear models were used to determine the efficacy of escitalopram in different patient subgroups. Furthermore, a factor analysis of the primary efficacy scale, the Liebowitz Social Anxiety Scale (LSAS), was undertaken, and a determination made of whether treatment effects were similar for the different symptom dimensions. Escitalopram was effective in both younger and older patients, in male and female patients, and in patients with more and less severe social anxiety symptoms. The LSAS factor analysis showed six factors, which were differentially associated with different areas of disability. Escitalopram was significantly superior to placebo for all six symptom dimensions. The treatment effects of escitalopram were independent of gender, symptom severity and chronicity, and comorbid depressive symptoms. A six-factor model of social anxiety symptoms is supported by the distinctive association between these symptom dimensions and different areas of disability, but did not predict differential response to escitalopram.