The differential effect of 2 types of interpolated interviews upon the perceptualmotor performance (Stein's Symbol-Gestalt Test) of brain-damaged and neurotic patients was investigated in 2 experiments. Cross-validated results indicate that Ss given disinterested, impersonal, irrelevant information-gathering interviews have a significantly lower rate of improvement over 3 successive administrations of the test than Ss given warm, supportive, anxiety-reducing interviews. Interaction effects were significant in the brain-damaged groups but not in the neurotics. The results support Goldstein's contention that brain damage results in a heightened dependency upon external cues and feedback and emphasize the importance of E variables in the clinical assessment of, and research, with brain-damaged populations.Emotional reactions and personality changes associated with brain damage have been described by many investigators (Brosin, 1952;Cameron, 1947;Goldstein, 1959;Reitan, 1955). The question of the degree to which such concomitants of cerebral dysfunction contribute to deficit in performance on intellectual and perceptual-motor tasks in the brain damaged has important implications for neuropsychological theories of behavior and, at a more practical level, for decisions about rehabilitation efforts, competency to handle business affairs, hospitalization, etc.A commonly reported emotional reaction to brain damage is increased anxiety. The deleterious effect of this affect upon performance in normal subjects (5s) on a variety of complex tasks has been well established (Sarason, 1960). There are two studies in 1 The first experiment was conducted while the authors were at Duke University Medical Center and represents partial fulfillment of the internship research requirements for the second author under the supervision of the first author. The second study was conducted at the Veterans Administration Hospital of the University of Oklahoma Medical Center; thanks are due to the staff and patients for their cooperation.