Although play and toys have been shown in the literature to benefit children in a variety of domains, relatively little is known about the effects of specific types of toys on the behavior and attitudes of children. Chapter 1 consolidates research that has attempted to discover the relationship between children's attitudes/behaviors and war toys. Research concerning the benefits of play and toys to children is presented, followed by the opposing viewpoints of aggression theorists regarding this issue. Research supporting each side of the debate is presented. Finally, variables that are likely to affect the war toy-child aggression relationship are discussed. In Chapter 2, an experimental study concerning the effects of violent toys on the aggressive play behavior of preschool children is presented. Twelve groups of three children, aged 3 to 5 years, participated in play sessions in which they were exposed to violent and nonviolent toys in a repeated measures design. The videotaped sessions were coded for: (a) children's aggressive behaviors and (b) the real or pretend play context in which aggression occurred. Analyses of variance revealed a significant main effect of toy condition for the observational data. Real, pretend, and total aggression occurred more often in play with violent toys than in play with nonviolent toys. Correlational analyses were performed on the survey data of participants and their parents, and between survey and observational data. Boys preferred more violent toys and television programs than girls; they also possessed more toy guns and played with them more frequently than girls. Children who preferred more violent toys and television shows had parents who possessed more positive attitudes toward spanking than parents of those children with less violent preferences, although no direct relationship between parental attitudes toward spanking and children's aggressive behaviors was identified. Finally, children of fathers with higher demands for mature behavior and greater use of firm-responsive control were found to exhibit more pretend aggression in play with nonviolent toys than children of fathers with lower maturity demands and less use of firm-responsive control. Berkowitz's aggressive cue hypothesis (1962, 1964, 1969, 1974, 1984) is generally supported.