The purpose of this study was to examine the relationships between child’s temperament and behavior problem according to child’s sex. The subjects for this study were children’s mothers of 2nd grade selected from elementary schools in Busan. Han, Sae-Young’s questionnaires(1996) on child’s temperament and Kim, Min-Kyung’s questionnaires(1998) on child’s behavior problem were used. Statistical techniques such as t-test, Pearson’s correlation, multiple regression, canonical were used. The results were as follows : (1) According to sex, there were no significant differences in child’s temperament. But there were significant differences in child’s behavior problem. The boys got higher scores than girls in hyperactive/distractive behavior and in hostile/aggressive behavior. (2) There were significant correlations between child’s emotionality and hyperactive/distractive behavior, between child’s emotionality and anxious/fearful behavior, between child’s emotionality and hostile/ aggressive behavior, and between child’s activity and hyperactive/distractive behavior. (3) The predicted variables for child’s hyperactive/distractive behavior were emotionality and activity. The predicted variable for chid’s anxious/fearful behavior and hostile/aggressive behavior was emotionality. (4) Among three dimensions of boy’s temperament, the effective variable was emotionality. And among three dimensions of boy’s behavior problem were anxious/fearful behavior and hostile/aggressive behavior. Among three dimensions of girl’s temperament, the effective variable was emotionality. And among three dimensions of girl’s problem behavior was anxious/fearful behavior. Implications of this study were discussed in terms of child’s temperament and behavior problem.