This is the first reported controlled trial of a partial benzodiazepine agonist, abecarnil, utilized in the treatment of generalized anxiety disorder (GAD). It was a sequential dose-finding study comparing 15-30 mg/day, 7.5-15 mg/day, and 3-9 mg/day to placebo for 3 weeks of treatment followed by abrupt discontinuation through placebo substitution. Although the two higher dose groups had high incidence of central nervous system (CNS) sedative adverse effects, the 3-9 mg/day group tolerated the medication well with no dropouts. The 3-9 mg/day group, in comparison to the two higher doses and placebo, demonstrated efficacy in global improvement ratings and Hamilton Anxiety Scale (HAM-A) scores. At Week 3, 61 percent of the abecarnil 3-9 mg/day group was rated as at least 50 percent improved on the HAM-A, compared to 30 percent of the placebo group. With abrupt discontinuation there were mild to moderate withdrawal symptoms and loss of efficacy in the two higher dose groups. However, in the 3-9 mg/day abecarnil group, there were few withdrawal symptoms and almost no loss of efficacy following discontinuation.