Using a nationally representative sample of 5,150 Dutch students who have been followed over a 6-year period, the presence of the Matthew effect was investigated for general language skills. The analyses do not reveal unmistakable evidence for the supposition that the rich get richer and the poor poorer. On the contrary, in schools with low starting levels students make more progress than in schools with higher starting levels. On the other hand, the analyses do show a widening gap between students with well educated and poorly educated parents. The gap between delayed and accelerated students was found to increase as well, but the initial disadvantage of boys was found to disappear.