Although co-occurring disorders are common in pathological gambling (PG), investigations of the response to pharmacotherapy in individuals with PG and co-occurring psychiatric symptomatology are limited. Thirteen subjects with DSM-IV PG and co-occurring anxiety were treated in a 12-week open-label trial of escitalopram. Subjects were assessed with the Yale-Brown Obsessive Compulsive Scale Modified for Pathological Gambling (PG-YBOCS; primary outcome measure), the Hamilton Anxiety Rating Scale (HAM-A), the Clinical Global Impressions scale (CGI), and measures of psychosocial functioning and quality of life. Those subjects who 'responded' (defined as a 30% or greater reduction in PG-YBOCS total score at endpoint) were offered inclusion in an 8-week double-blind discontinuation phase. PG-YBOCS scores decreased from a mean of 22.2+/-4.5 at baseline to 11.9+/-10.7 at endpoint (P=0.002) and 61.5% were responders. Scores on the HAM-A decreased by 82.8% over the 12-week period (mean of 15.9+/-3.2 at baseline to a mean of 2.8+/-3.6 at endpoint) (P<0.001). On the CGI, 38.5% of subjects (n=5) were 'very much improved' and 23.1% (n=3) were 'much improved' by study endpoint. The Sheehan Disability Scale, Perceive Stress Scale and Quality of Life Inventory all showed improvement (P< or = 0.001, P=0.002 and P=0.029, respectively). The mean end-of-study dose of escitalopram was 25.4+/-6.6 mg/day. Of three subjects assigned to escitalopram during the discontinuation phase, none reported statistically significant worsening of gambling symptoms. However, one subject assigned to placebo reported that gambling symptoms returned within 4 weeks. Open-label escitalopram treatment was associated with improvements in gambling and anxiety symptoms and measures of psychosocial functioning and quality of life. Larger, longer, placebo-controlled, double-blind studies are needed to evaluate further the safety and tolerability of escitalopram in the treatment of PG and co-occurring anxiety.