Using a decision tree algorithm, we set up a model for the children's behavior self-regulation. Children aged 3 to 7 years from kindergarten to the first grade of primary school (Mage =5.7; N=299) were investigated in this study. First, children's behavioral self-regulation was measured using the head toe knee shoulder (htks) task. Secondly, the data attributes of children and their families were collected based on the questionnaire survey. Thirdly, the Gini decision tree method was used to construct the classification model of children's behavior self-regulation. The result of the Gini decision tree classification model presents that 1) children's self-regulation increases rapidly with age, 2) the self-regulation of girls is slightly higher than that of boys, and 3) children of the same age group who go to school early have higher self-regulation. The findings highlight developmental differences in early childhood behavior self-regulation and provide a rationale for future interventions.