The objective of the study was to evaluate the response of generalized anxiety disorder (GAD) patients with prominent gastrointestinal (GI) symptoms to pregabalin (PGB) treatment. Data were pooled from six double-blind, placebo (PBO)-controlled, 4-6 week trials in outpatients who met Diagnostic and Statistical Manual of Mental Disorders, 4th edition criteria for GAD with a minimum Hamilton Anxiety Rating Scale (HAM-A) total score of 20. Treatment response was evaluated for three PGB fixed-dosage groups: 150, 300-450, and 600 mg/day, and for fixed doses of a benzodiazepine (alprazolam, 1.5 mg/day; lorazepam, 6 mg/day). A GI-high subgroup (high GI symptomatology) was defined by a baseline HAM-A item-11 (GI) score of 3 or greater (severe/very severe). At baseline, 301 patients (16.2%) met criteria for the GI-high subgroup. Baseline characteristics were approximately similar for the four study treatments in the GI-high subgroup. For the GI-high subgroup, last observation carried forward (LOCF) endpoint reduction in HAM-A was significantly higher on PGB-300/450 -13.8+/-1.2 and PGB-600 -14.7+/-1.0 compared with PBO -10.1+/-0.9 (P<0.01 for both comparisons); but the difference on PGB-150 did not achieve significance (-13.5+/-1.6; P=0.083). Also in the GI-high subgroup, endpoint reduction in HAM-A item-11 was significantly higher on PGB-300/450 compared with PBO (-1.93+/-0.16 vs. -1.52+/-0.13; P=0.04), but did not achieve significance on PGB-600 mg (-1.89+/-0.14; P=0.06), or PGB-150 mg (-1.90+/-0.23; P=0.16). In the GI-high subgroup, treatment with a benzodiazepine was not associated with significant endpoint reduction in either the HAM-A total score or the HAM-A item-11 score. Patients in the GI-high subgroup had higher discontinuation rates when treated with benzodiazepines, whereas treatment with PGB 300-600 mg/day was not associated with treatment-emergent worsening in GI symptoms compared with placebo. Treatment with PGB improved overall levels of anxiety, as well as specifically improving GI symptoms.