According to Reber (1992), implicit learning should vary only minimally with age and intelligence quotient (IQ), whereas explicit learning should depend heavily on both age and IQ. To evaluate these predictions, children from two age groups (5–7 yean and 10–12 years) were selected to form low, medium, and high IQ subgroups. They were then tested on an implicit and an explicit learning task. The implicit task, adapted from Lewicki (1986), used a number of 4 × 4 matrices which displayed 16 pictures in different arrangements. One picture was of a house, and the children were required to learn the incidental covariation of (a) the experimenter's side of approach (left or right), with the house positioned in the upper or lower half of the matrix; and (b) the colour of the matrix board and cover (red or blue), with the house positioned in the left or right half of die matrix. The explicit task had a similar logical structure to the implicit task, but (a) was implemented using different stimulus dimensions; (b) was presented so that the stimulus dimensions were not only salient, but were also explicitly identified to the children; and (c) used instructions that identified the nature of the rules to be learned, ensuring that learning was intentional. The reported results are inconsistent with the claim that age differences exist only in explicit learning, since implicit learning also improved significandy with age. However, the results are consistent with the claim mat implicit learning is invariant of, whereas explicit learning covaries with, IQ. The study is discussed in the context of the applicability of the implicit‐explicit distinction, and like distinctions, to theories of cognitive deve