Indirect evidence, mostly pharmacologic, has suggested a role for brain neurotransmitter amines such as norepinephrine in the production of depression or mania. Clinical investigations have supported this concept but also indicate that depression is probably a biochemically heterogeneous group of illnesses. There may be a clinically, biochemically, and pharmacologically definable subtype of depression in which there is a disorder of norepinephrine metabolism or disposition in brain. I review here the experimental data from which this hypothesis is derived.