This investigation--the first Dutch longitudinal research project on reading and spelling spanning the whole 6-year period of primary education--examined (1) whether it is possible to identify certain aspects of literacy--namely decoding skills, reading comprehension, and spelling skills--within verbal intelligence as latent variables, and whether, if this is the case, these variables remain sufficiently constant during the elementary school; (2) the interrelations among these three skills in terms of growth; and (3) whether causal effects found in a cross sectional approach would be "explained away" in longitudinal research. Subjects, 582 first-graders from 12 randomly selected Dutch schools, were tracked in grades one, two, three, and six, beginning in 1979. Results indicated that in grade three the influence of reading comprehension on spelling was lower than in the first two grades, but that in grade six this influence increased considerably. The direct influence of decoding skills on spelling, however, was smaller than in the first three grades. Results also showed that in grade six correlations and path coefficients between decoding skills on the one hand, and reading comprehension and spelling on the other, remained as a remnant of a causal connection that existed during an earlier developmental stage, but that these relationships were developmentally limited. In most cases the connections between the latent variables of the same name were much stronger than the relations with the other ones, an indication that the differences related to these latent variables were stable, especially for decoding skills. (Twenty-one references, two tables of data, and three figures are included.) (NKA)