The prosocial behavior of 52 preschool children was assessed using three different approaches: naturalistic observation, laboratory measures, and teacher ratings. During the naturalistic observation, an observer, either male or female, focused on one child at a time and recorded any examples of prosocial behavior demonstrated by the child as well as the antecedents (whether a request preceded the behavior and, if so, which type of request) and the consequences of this behavior (whether the recipient expressed gratitude or reciprocated). Each child was observed for an average of 79 minutes over a period of 5 months. During the last month of the observation, laboratory measures of perspective taking, prosocial behavior, and empathy were taken. Three teachers then rated the children's prosocial behavior under different eliciting situations: explicit request from the teacher, explicit request from another child, or spontaneous behavior without a request. Analysis of the antecedents and consequences of the prosocial behavior, the structured measures of perspective taking, and the components of empathy suggest the relative importance of various contextual and motivational influences on each type of prosocial behavior. Preschool children demonstrate a sensitivity to the needs and feelings of their peers, and this capacity, though not tapped by traditional measures of perspective taking, may mediate prosocial behavior in the natural setting. Methodological implications are also presented. (Author/MP)