This study attempts to explore the nautre of the Korean parents' demand for their children's education, and common features and differences in the nature between social classes. The data analyzed are collected by interviewing the 96 couples of students and their parents on the one hand, and by conducting the questionaire survey of the 4,000 couples of students and their parents on the other hand. Among the surveyed students and parents, 1,390 couples are finally used in this analysis. Crosstabulation and logistic regression analysis statistical methods are employed. Most of parents in Korea, regardless of their social classes, desire of their children to have higher education. However, when we carefully look at the types and the level of higher education expected by parents, we find differences between social classes. The higher the parents' educational level is, the stronger the tendency to expect their children to go to 4-year colleges and universities(in comparison to 2-year junior colleges) and graduate schools(in comparison to colleges and universities). Middle class parents are more likely to expect their children to go to 4-year colleges and universities(in comparison to junior colleges) and graduate schools(in comparison to colleges and universities), than are working class and urban marginal class parents. The percentage of parents expecting their children to go to graduate schools is 60.4% among the families in which mother has 4-year college or more education, whereas it is only 11.8% among those families in which mother has primary or less education. On the contrary, the percentage of parents whose expectation level is the graduation of junior colleges is 4% for the former families, whereas it is 21.4% for the latter families. We can see the social class differencies even when we control for the children's school achievement. These results show us that social class differences in the parental expectation for their children's education are still existed in Korea in the feature different from that used be. Further, the results suggest that upper middle class parents with higher education induce “educational inflation” in the next generation. The results of this study also present that parents and their children are not necessarily agreeded among themselves on the educational level children will reach in the end. Parents tend to have the higher expectation level than their children to have. The disagreement appears more strongly among families with highly educated parents or middle class families, which means that middle class children more suffer from the parental pressure for improving their academic achievement than do working class or urban marginal class children. The findings also show that Korean parents' supports for their children's education are mainly geared to the better academic achievement, in preparation for the college entrance examination. Among private extracurricular lessons primary school children take, those lessons helping school academic work have a majority. It is a matter of course that the private lessons of middle school children are more concentrated on those helping school academic work. The supports for developing their children into well-rounded persons are minor even in the primary school level. When we make a comparison of the pattern of children's private lessons between social classes, little significant differnces are found between social classes for the primary school children, except that middle class children tend to take a little more private lessons promoting the development of well-round personalities than do working class or urban marginal class children. However, as for middle school children, the former tend to take far more private lessons helping school academic work than do the latter. The results show that opportunities for developing all-round personalities are scarce even in the primary school level regardless of social classes.