The present meta-analysis assessed concurrent and longitudinal associations between parental educational expectations and child achievement, and factors that mediate the effect of expectations on achievement. A systematic search in electronic databases identified 169 studies that were included in a random-effects meta-analysis. We found small to moderate bivariate cross-sectional (r = .30) and longitudinal associations (r = .28) between parental expectation and achievement which persisted after statistically controlling for socioeconomic status. Associations varied, in part, by children's age, socioeconomic status, ethnicity, matching of type of expectations and achievement, type of expectation assessed, publication status, and informant. The analysis of cross-lagged effects indicated that parental expectations predicted change in child achievement, thus indicating that expectations had an effect over and above the effect of prior achievement. Effects of expectations on change in achievement were even stronger (r = .15) than the effects of achievement on change in expectation (r = .09). Parental expectations tended to be higher than the child achievement. Associations between expectations and achievement were partially mediated by educational expectations in the offspring, child academic engagement, and academic self-concept, and to a lesser extent, by parental achievement-supportive behaviors. We conclude that parents are recommended to communicate positive educational expectations to their children. The transmission of positive expectations to the offspring and the encouragement of academic engagement seem to be more effective in realizing parental expectations than parental behavioral academic involvement such as checking homework and staying in contact with teachers.