Depressed patients given a loading dose of lithium on the first of 2 successive days of partial sleep deprivation (PSD), and kept at maintenance levels thereafter, showed significantly greater prolongation of the antidepressant effects of PSD than patients treated with PSD and placebo, even though the acute elevation in mood derived from PSD was as great on placebo as on lithium. Depression was assessed 3 days after PSD with an augmented version of the Hamilton Rating Scale for Depression. Patients on lithium alone, without PSD, did not have the acute elevation in mood seen in the two PSD groups and had significantly less improvement in depression than those who received PSD with lithium.