This article examines how the alternatives of scapegoating or social-emotional leadership are determined in small discussion groups. The hypothesis asserts that the alternative “selected” by the group depends on whether the task leader supports the low-status member. An experiment designed to test this hypothesis indicated that scapegoating of the low-status member occurs when the task leader is somewhat hostile toward him. However, social-emotional leadership was not more in evidence when the task leader supported the low-status member. Instead, such support had the unanticipated consequence of increasing the task leader's legitimacy in performing task activity, thus decreasing both scapegoating and social-emotional leadership.