BACKGROUND: This study compared the efficacy and tolerability of escitalopram, a newer SSRI, with paroxetine in the treatment of generalized anxiety disorder (GAD). METHODS: Patients with DSM-IV-defined GAD were randomized to receive 24 weeks of double-blind flexible-dose treatment with either escitalopram (10-20 mg/day) or paroxetine (20-50 mg/day), followed by a 2-week, double-blind, down-titration period. Mean change from baseline to endpoint (LOCF) in Hamilton Anxiety Scale (HAMA) scores was the primary efficacy variable. RESULTS: Mean baseline HAMA scores for the escitalopram (N = 60) and paroxetine (N = 61) groups were 23.7 and 23.4, respectively. After 24 weeks of treatment, mean changes in HAMA scores were -15.3 and -13.3 for escitalopram and paroxetine, respectively (p = 0.13). Significantly fewer patients withdrew from escitalopram than paroxetine treatment due to adverse events (6.6% vs. 22.6%; p = 0.02). The frequency of treatment-emergent adverse events was higher with paroxetine vs. escitalopram: overall (88.7% vs. 77.0%), insomnia (25.8% vs. 14.8%), constipation (14.5% vs. 1.6%), ejaculation disorder (30.0% vs. 14.8%), anorgasmia (26.2% vs. 5.9%), and decreased libido (22.6% vs. 4.9%). Conversely, diarrhea and upper respiratory tract infection were reported more with escitalopram than paroxetine (21.3% vs. 8.1%, and 14.8% vs. 4.8%, respectively). CONCLUSIONS: These results support the use of escitalopram as a first-line treatment for GAD.