This article presents the results of a lexical decision experiment in which the base frequency (BF) effect is explored in reading disabled children and skilled readers. Three groups of participants were created. The first group was composed of children with reading disorders, the second group of skilled readers matched with the first group for chronological age and the third group of skilled readers matched for vocabulary size. The results of the experiment showed strong effects for Group, BF and also for the Group by BF interaction. Children matched for chronological age with children with reading disorders were significantly faster and more accurate than children of the other groups, who did not show any difference from each other. The effect of BF showed that children responded faster to stimuli composed of frequent bases than to stimuli with less frequent bases. However, the analysis of the interaction between Group and BF showed that only the skilled readers matched to children with reading disorders for chronological age benefited from the BF effect. The results of the experiment are discussed in the framework of theoretical accounts of morphological processing in children as well as considering the role played by the experimental task. (Contains 3 tables and 1 figure.)