This paper attempts to show that severely stressful conditions can lead to a deficit in avoidance-escape responding by reducing noradrenergic activity in the brain. It is argued that this stress-induced neurochemical change explains more adequately an avoidance-excape deficit seen previously in dogs after severe inescapable shock than does the "learned helplessness" explanation originally offered for this effect. A series of six experiments are described that test the stress-neurochemical change explanation, which is called the "motor activation deficit" hypothesis. The first experiment showed that a brief exposure to cold swim, which reduced central noradrenergic activity but did not induce "learned helplessness," produced an avoidance-escape deficit similar to that observed after inescapable shock in the original "learned helplessness" studies. Subsequent experiments demonstrated that the deficit induced by cold swim was also similar to the original deficit in that it too dissipated with the passage of time and was reduced by pretraining animals in the correct response. Further experiments then showed that the deficit induced by cold swim was (a) aggravated by raising the height of the barrier between compartments in the shuttle box, and (b) did not occur if the avoidance-escape task used required little motor activity, both of the findings being predicted by the "motor activation deficit" hypothesis. Finally, an avoidance-escape deficit was produced by inescapable shock, and it was found that this stressor likewise did not impair avoidance-escape responding if the task required little skeletal activity, a result not predicted by the "learned helplessness" hypothesis. The paper concludes by discussing how the results presented in this paper, as well as in succeeding papers, support the "motor activation deficit" explanation for certain avoidance-escape deficits, defining the cases to which the explanation applies and discussing the neurotransmitters involved.