Few pharmacotherapy trials have been undertaken in young people with anxiety disorders. Of those conducted, few are placebo-controlled or blinded, and often sample size is small, making interpretation of the data difficult. Case report and uncontrolled trial data generally support the efficacy of pharmacotherapy in many of the anxiety disorders seen in young people. However, most attempts to confirm these impressions in controlled trials have not been as encouraging. Further controlled studies in larger, diagnostically homogeneous samples are needed. At present, the decision as to whether or not to use medication in this patient population must be made on clinical grounds. Evidence is accumulating to suggest that anxiety disorders in young people can become chronic and cause at least moderate functional impairment in some individuals. This possibility has to be weighed against the potential for adverse effects of many of the drugs used clinically. Serious adverse effects appear only in a minority of patients; however, there does not seem to be a reliable method of predicting which patients might be at risk. Benzodiazepines are used to treat anxious children, notwithstanding concerns about dependence, behavioural disinhibition, cognitive impairment and mood changes. The tricyclic antidepressants (TCAs) are generally well tolerated, but their effects on cardiac conduction at higher plasma concentrations are well documented and there have been sporadic reports of sudden death associated with their use in children. Toxicity in overdose is an added concern. The use of selective serotonin (5-hydroxytryptamine; 5-HT) reuptake inhibitors in children and adolescents is widespread, but few rigorous supportive trials have been conducted. The occurrence of both potentially serious adverse effects and other less serious, but troublesome, adverse effects and a possible discontinuation syndrome may complicate the use of these agents in younger patients. However, their relative safety in overdose and the apparent reversibility of adverse effects may favour their use over the TCAs. Other drugs used in this setting include buspirone and beta-blockers. No controlled trials are available at this time, but their adverse effect profiles appear to be relatively favourable and further study is warranted, given the promising nature of the uncontrolled data. The reversible monoamine oxidase inhibitors may be useful in younger anxious patients, but controlled data are needed. The decision to use pharmacotherapy should be made after consideration of multiple factors, and firm recommendations for the use of drug therapy in anxiety disorders in children and adolescents await more rigorous data.