Resistance to antidepressant therapy is a common clinical problem in the treatment of affective disorders. Adjunctive low dose lithium is a promising strategy based on biochemical models and encouraging clinical trials. After a mean duration of 9.2 months of conventional therapy, 16 healthy patients with treatment-resistant depression were treated for a minimum of 2 weeks with either adjunctive lithium or placebo using a double-blind design. We found no difference between the two groups in rate or degree of response. The two most dramatic responses occurred in patients treated with placebo, although 50% of patients treated with lithium had at least a partial response. The number of patients studied was clearly inadequate to avoid a type 2 error. The cumulative response rate reported in the literature of greater than 60%, however, suggests that lithium is indeed an effective adjunct in some patients with treatment-resistant depression. Our patients differed from those in other studies in that they were treated with a lower dose of lithium, the duration of conventional antidepressant therapy was longer, and, finally, they were less depressed and possibly depressed for a longer period. These differences may explain the comparable lithium and placebo responses in this study.