A classic issue in educational research with considerable theoretical, methodological and practical implications concerns the relative amount of influence of schooling and age, respectively, on the development of intellectual performance. Previous studies, which suffer from different kinds of methodological limitations, indicate a considerable effect of schooling. Two problems, which both might imply an overestimation of the schooling effect, are selection effects related to age differences, i.e. students who are under-/over-aged in relation to grade level, and the assumption of linearity of the age effect. The main purpose of the project is to gain further knowledge about these phenomena by the use of more powerful methods than previously used and taking advantage of unique longitudinal data. The specific purposes are to study: (1) the effects of selection effects on the estimated regression of achievement on age and the linearity assumption, by the use of population data for successive cohorts; (2) the homogeneity of the regression on age, and control for selection effects, by the use of missing data modeling; and (3) the relative amount of influence of age and schooling, and differential effects of schooling on differential aspects of intellectual performance, by the use of a design relying on simultaneous variation in age and amount of schooling based on data from tests used for enlistment to military services.