OBJECTIVES: To assess the cost-effectiveness of escitalopram versus generic paroxetine in the treatment of generalized anxiety disorder (GAD) in Norway. METHODS: Cost-effectiveness analysis was based on a decision-tree model, reflecting the current guidance for the treatment of GAD. Escitalopram 10-20mg/day was compared to paroxetine 20-50mg/day, the only SSRIs licensed for the treatment of GAD in Norway. Initial treatment started in primary care; failure to respond to a second-line SSRI after switch from initial treatment resulted in referral to a psychiatrist. The evaluation was performed over a 9-month time horizon. Model data sources included published clinical trials and standard national data sources. The main effectiveness outcome was first-line treatment success, defined as a response after 12 weeks of treatment (³50% reduction in HAM-A) and absence of relapse during the following 24 weeks. Other outcomes included quality-adjusted life years (QALYs), rate of maintained treatment response (response at week 12 maintained at week 36) and rate of referrals to secondary care. The analysis was performed from the societal perspective and evaluated direct treatment costs, including out-of-pocket payments and costs covered by health insurance. RESULTS: Patients treated with escitalopram had 11.3% higher first-line success and 5.4% higher maintained response rates compared with paroxetine-treated patients. Rates of referrals to secondary care were 5.5% lower with escitalopram compared with paroxetine. Per patient direct treatment costs associated with escitalopram and paroxetine treatment amounted to NOK 4424 and NOK 4172, respectively. Incremental cost-effectiveness ratio (ICER) was NOK 2,230 per first-line treatment response and NOK 37,612 per QALY. Sensitivity analyses demonstrated robustness of the model to changes in key input parameters. CONCLUSION: Escitalopram appears to be cost-effective compared with paroxetine in the treatment of GAD in Norway, based on direct treatment costs