A meta-analytic review of the efficacy of pharmacological treatment in generalized anxiety disorder was conducted. The main substance classes were compared: benzodiazepines and azapirones. The impact of methodological variables was investigated such as sample size and use of a placebo run-in. After a comprehensive literature search to May 2002 (via databases, hand search, secondary sources, internet, contact of researchers, and pharmaceutical companies), the results of 48 studies were integrated. Weighted Hedges g was computed and a random-effects analysis was done. Effect sizes were computed for anxiety, depression, and clinical significance. Sensitivity analyses were conducted. Pharmacotherapy was superior to placebo in all symptom categories. Azapirones and benzodiazepines were equally effective. Compliance (as measured by dropout rate) was higher for benzodiazepines. Only sample size was significantly associated with effect size. Pharmacotherapy, especially benzodiazepines and azapirones, is effective in the short-term treatment of patients with generalized anxiety disorder. There was no superiority of 1 drug class in reducing symptomatology.