This paper reviews the functional brain imaging studies in depressed patients treated with sleep deprivation. Sleep deprivation is an excellent experimental model of antidepressant treatments which offer new opportunities to understand the basic neural mechanisms. Its antidepressant effects are efficacious and rapid; sleep deprivation is easy to administer, inexpensive, and relatively safe; it can be studied in patients, normal controls, and animals; and it may lead to new treatments and new paradigms for antidepressant therapies. Seven published papers, coming from five different research centers, using either positron emission tomography (PET) with 18fluorodeoxyglucose (FDG) or single photon emission computerized tomography (SPECT) with Technetium-99-bexamethyl propyleneamine oxime (HMPAO) have relatively consistent findings. First, before sleep deprivation, responders have significantly elevated metabolism compared with non-responders, and usually the normal controls, in the orbital medial prefrontal cortex, and especially in the ventral portions of the anterior cingulate cortex. Secondly, after sleep deprivation, these hyperactive areas normalize in the responders. The magnitude of the clinical improvement was significantly correlated with decreased local glucose metabolic rate or cerebral blood flow in three studies. The results are consistent with some but not all functional brain imaging studies of antidepressant medications in depressed patients. Finally, a SPECT study using a radioactively labeled D2 receptor antagonist suggests that the antidepressant benefits of sleep deprivation are correlated with endogenous release of dopamine.