The initial hypothesis was that physically disabled help-seekers will receive more aid from normal people than will help-seekers when both types of actors display positive personal qualities, whereas the reverse outcome will occur when both display negative qualities. In the first experiment a wheelchair-bound or normal confederate was either friendly and achievement oriented (positive condition) or caustic and apathetic (negative condition) while administering verbal tasks to subjects who were later asked to help the confederate. Contrary to prediction, subjects offered more help to the normal than to the disabled tester in the positive condition, and showed the reverse pattern in the negative condition. It was conjectured that subjects in the positive condition were annoyed by the disabled person's display of “normal” characteristics, whereas in the negative condition they sympathetically accepted the disabled person's inadequacies as befitting a victim of severe misfortune. A new experiment was done, which replicated the independent variables of the first study, but substituted two new dependent variables: covert anger arousal and perception of the confederate as happy-unhappy. As expected, covert anger was greater in the wheelchair-positive condition than in the normal-positive condition, and greater in the normal-negative condition than in the wheelchair-negative condition. Regardless of the confederate's physical status, subjects perceived negative behavior as a sign of unhappiness.