The aim of this study was to design a model of relational and overt aggression in children based on: parenting stress and maternal parental self-efficacy mediated by family communication patterns. Participants were 540 people (275 boys and 265 girls) from 9 to 11 year old primary school students in Tehran in the academic year 1400-1399 with their mothers who participated in the study by available sampling method. Participants answered questionnaires of relational and overt aggression in preschool children, parenting stress, Domka parental self-efficacy, Ritchie and Fitzpatrick family communication patterns. Spss software version 26, Amos software-24 and R software were used for statistical analysis of the data. The findings indicated a good fit of the collected data with the designed model. Parenting stress has a significant role in the relationship and overt aggression of children. Parenting stress has a significant direct effect on family communication patterns. Parental self-efficacy does not play a direct role in family communication patterns. The variable of parental self-efficacy with the effect of 0.04 has no significant role on the variable of aggression and overt in children. Family communication patterns play a direct role in relation to the explicit relationship variable of children. Parenting stress through family communication patterns has a significant role on the relational and overt aggression variable. Self-efficacy through communication patterns does not play a significant role in relationship and overt aggression. The findings of the present study are a step towards developing predictive theoretical models for relational and overt aggression in children.