The authors reviewed pharmacologic treatment of generalized anxiety disorder (GAD), particularly treatment with benzodiazepine agents, and compared the antianxiety effects and dependence risks of these agents with those of nonGABAergic compounds such as buspirone--a new psychotropic drug--in the treatment of chronic anxiety. Forty outpatients who met DSM-III criteria for GAD were randomly assigned to double-blind treatment with buspirone or lorazepam. After 8 weeks, treatment was abruptly stopped and withdrawal reactions were evaluated at Weeks 9 and 10. After 3 to 4 weeks, buspirone's efficacy in treating GAD symptomatology was found to be comparable with lorazepam's, except for sleep disturbances, which were minimally affected by buspirone. After treatment was discontinued, buspirone-treated patients were free from withdrawal symptoms, while the symptoms of lorazepam-treated patients worsened at Week 9.