Objectives: Depressive disorders are persistent, recurring illnesses that impose enormous personal suffering on individuals and their families. Major depression alone is estimated as the fourth most important cause of worldwide loss in disability-adjusted life years and is likely to become the second most important within 20 years. A continued quest for more effective treatments has spawned newer antidepressants and herbal treatments, which have contributed to explosive growth in antidepressant prescribing, increasing pharmacy costs, and wider but sometimes confusing choices for clinicians and patients. This evidence report provides a comprehensive evaluation of the benefits and adverse effects of newer pharmacotherapies and herbal treatments for depressive disorders in adults and children. Search Strategy: Pertinent literature from 1980 to January 1998 was identified from a specialized registry of controlled trials, meta-analyses, and experts. The registry contained trials addressing depression that had been identified from multiple electronic bibliographic databases, hand searches of journals, and pharmaceutical companies. The search, which yielded 1,277 records, combined terms 'depression,' 'depressive disorder,' or 'dysthymic disorder' with a list of 32 specific 'newer' antidepressant and herbal treatments. Selection Criteria: Randomized controlled trials were reviewed if they (1) were at least 6 weeks in duration; (2) compared a 'newer' antidepressant with another antidepressant (newer or older), placebo, or psychosocial intervention; (3) involved participants with depressive disorders; and (4) had a clinical outcome. Two or more independent reviewers identified 315 trials that met these criteria. Data Collection and Analysis: Two persons independently abstracted data from each trial. Data were synthesized descriptively, paying attention to participant and diagnostic descriptors, intervention characteristics, study designs and clinical outcomes. Some data were analyzed quantitatively using an empirical Bayes random-effects estimator method. Primary outcomes were response rate, total discontinuation rates (dropouts), and discontinuation rates due to adverse events. Response rates were defined as a 50 percent or greater improvement in symptoms as assessed by a depression symptoms rating scale or a rating of much or very much improved as assessed by a global assessment method. Main Results: There were 264 trials that evaluated antidepressants in patients (adults and children) with major depression. Of these, 81 compared newer agents with placebo, 150 newer with older agents, 32 newer agents with newer agents, and 1 newer agent with psychotherapy. There were 14 trials evaluating hypericum (St. John's wort), 27 trials each in primary care patients and older adults, 10 trials limited to patients with specific concomitant illnesses, 9 trials in patients with dysthymia, 3 trials each in patients with mixed anxiety depression and subsyndromal depression, 2 trials in adolescents, and 1 in the postpartum setting. Most trials were conducted in outpatients and examined only acute phase treatment of less than 12 weeks' duration. Newer antidepressants were more effective than placebo in treating major depression (risk ratio 1.6, 95% CI 1.5 to 1.7) and dysthymia (risk ratio 1.7, 95% CI 1.3 to 2.3). They were effective among older adults and in primary care patients. In general, there were no significant differences in efficacy among individual newer agents or between newer and older agents. Hypericum (St. John's wort) was more effective than placebo in treating mild to moderately severe depressive disorders (risk ratio 1.9, 95% CI 1.2 to 2.8). Whether hypericum (St. John's wort) is as effective as standard antidepressant agents given in adequate doses was not established. No significant differences were found between newer and older antidepressants in overall discontinuation rates. Selective serotonin reuptake inhibitors (SSRIs), reversible inhibitors of monoamine oxidase A (RIMAs) and hypericum (St. John's wort) had fewer dropouts due to adverse effects than first generation tricyclic agents (TCAs). When compared with first generation TCAs, SSRIs had higher rates of diarrhea, headache, insomnia, and nausea. SSRIs had lower rates of blurred vision, constipation, dizziness, dry mouth, tremors, and urinary disturbances. Conclusions: Newer antidepressants are clearly effective in treating depressive disorders in a variety of settings. Multiple agents are effective. In general, there are no significant differences in efficacy between newer antidepressants and first and second generation TCAs nor between different classes of newer antidepressants. Newer antidepressants have similar overall discontinuation rates as older antidepressants but varying side-effect profiles. Some newer agents, such as RIMAs and SSRIs, have fewer dropout rates due to adverse events compared with first generation tricyclic agents. Available trial data have several limitations. Most trials (74%) lasted 6 to 7 weeks; 81% reported dropout rates of greater than 20 percent. Fewer than 5 percent reported health-related quality of life outcomes. The actual treatment environment within which trials were carried out was rarely described. Few trials addressed effectiveness, and little information exists for children and adolescents; refractory depression; depression co-occurring with comorbid illness; and specific disorders such as mixed anxiety depression, subsyndromal depression, and depression in the postpartum setting.