A questionnaire designed to assist in the classification of depressive illness, was administered to 20 acute depressive patients and 10 normal controls. Patient classification according to the decision rules employed by the questionnaire, were compared with patient classification arrived at by symptomatic (diagnostic) and physiological (G.S.R. inhibition threshold) means. It was found that a preponderance of patients classified as "endogenous depressives" by the questionnaire technique, also had clinical diagnoses of endogenous depression, and had low G.S.R. inhibition thresholds relative to normals. Furthermore a preponderance of patients classified as "non-endogenous depressives" by the questionnaire technique, had clinical diagnoses of neurotic depressive illness, and had high G.S.R. inhibition thresholds relative to normals. This supports the assertion that the questionnaire technique has some validity in the classification of depressive illness.