Based on concepts derived from the literature on aggression, this study developed a theoretical model that hypothesizes a positive correlation between preference for an escalation of aggression and capability, that both preference for escalation and capability have a positive effect on the escalation pattern, and that boys and younger students will tend to have a higher level of aggressiveness in the constructs of escalation than girls and older students. In testing this model, stratified sampling was used to obtain data from 799 students in the seventh, eighth, and ninth grades of junior high schools in a northern Israeli city and its suburbs. The instrument used in interviews with the youth included three main constructs derived from the model. One section of the interview instrument contained two groups of questions, one referring to preferences in reacting to verbal aggression and the other to preferences in reacting to physical aggression. A second construct involved a self-evaluation of the potential to act in a nonaggressive/aggressive manner verbally or physically. The third construct contained two groups of questions, one referring to the pattern of reacting to verbal aggression and the other to the pattern of reacting to the physical aggression. Operationalizing the interaction-based unit of analysis for both attitudes and behaviors included two components: measurement of the degrees of escalation and escalation tendency based on these measurements. Analysis with a structural equation model confirmed that there is a significant positive correlation between the constructs of the so-called escalation preference and capability; both significantly influenced the escalation pattern. In addition, boys and younger students apparently had a higher tendency toward escalating conflict than girls and older students. Suggestions are offered for future research in this area. 1 figure, 4 tables, and 40 references