As part of a double-blind clinical trial comparing phenelzine and amitriptyline in outpatients with predominantly major depressive disorder, plasma tricyclic antidepressant drug concentrations were measured in 83 amitriptyline-treated patients. In 29 of these patients, hydroxymetabolites were also assayed. Patients were treated for 6 weeks at a fixed dose of 150 mg/day of amitriptyline after the first 5 days. Therapeutic outcome was assessed with a structured depression interview schedule, the Symptom Checklist-90, a side effects checklist, and a global improvement scale. Steady state plasma levels of 10-hydroxynortriptyline were in the same range as amitriptyline or nortriptyline concentrations. Clinical response did not relate significantly to plasma levels of either the parent drug, its metabolites, or the sum of all four pharmacologically active substances. Minimum threshold tricyclic antidepressant levels for therapeutic effect were not found. Assay of its active hydroxymetabolites does not appear to improve the clinical utility of routine amitriptyline level monitoring in patients with major depression in an outpatient setting.