This investigation explored the belief that physical activity therapy has a neurological value as part of the total treatment of mental patients. Twenty-two adolescents hospitalized with depression were administered a battery of diagnostic tests. Relationships among the Braininks-Oseretsky Motor Proficiency test (12 subtest scores) and the Luria-Nebraska Neuropsychological Battery (14 subtest scores) were examined. All patients were nonmedicated and on a neurotransmitter controUed diet for 48 hours prior to testing. Results indicate that certain forms of motor performance can be predicted from measures indicative of both structure and fonction of brain behavior. The multiple R2 ranged from .80 to .06, with variance in balance, bilateral coordination, and fine motor skills being predicted from the receptive speech, tactile, right hemisphere, left hemisphere, expressive speech, motor, and rhythm measures of brain behavior. When explaining neuropsychological battery scores from motor proficiency scores, the R2 ranged from .65 to .28, with intelligence and expressive speech being predicted from the gross motor composite, upper limb speed, and dexterity, balance, and fine motor composite scores, respectively.