The objective was to examine duloxetine 60-120 mg/day treatment for relapse prevention in adults with generalized anxiety disorder (GAD). Adult patients (N=887; mean age=43.3 years; 61.0% female) with DSM-IV-TR-defined GAD diagnosis were treated with duloxetine for 26 weeks. Patients who completed open-label phase and were treatment responders (>or=50% reduction in Hamilton Anxiety Rating Scale total score to <or=11 and "much"/"very much improved" ratings for the last 2 visits of open-label phase) were randomly assigned to receive duloxetine or placebo for a 26-week double-blind continuation phase. Relapse was defined as >or=2-point increase in illness severity ratings or by discontinuation due to lack of efficacy. During the double-blind phase, placebo-treated patients (N=201) relapsed more frequently (41.8%) than duloxetine-treated patients (13.7%, N=204, P<or=0.001) and worsened on each outcome measure (P<or=0.001, all comparisons). Duloxetine 60-120 mg/day treatment was efficacious and reduced risk of relapse in patients with GAD.