The present research represents the final 2 years of a 5-year longitudinal investigation of (a) confrontation-based, word-retrieval deficits in dyslexic children; (b) the role of vocabulary development in these deficits; (c) the relationship between confrontation naming performance and three carefully defined aspects of reading performance in the general population and in eight dyslexic case studies; and (d) the possible specificity of word-retrieval deficits in dyslexia. Results indicate enduring problems in word-retrieval processes for dyslexic children across the primary grades and into middle childhood. Second, these deficits cannot be explained by simple vocabulary deficits. Third, these results in conjunction with our earlier data consolidate a pattern of differential relationships between specific reading and confrontation naming skills that are based on development and on the level of processes involved. Trends within case studies suggest the more pronounced the retrieval deficit, the more global the reading impairment. And fourth, there appear to be some specific differences in the basis of word-retrieval problems between dyslexic and garden-variety or lower achieving readers. Results are discussed within a speculative framework that implicates problems in timing as a possible predetermining condition in the dyslexias.