We tested the influence of list context on word frequency and length effects on the reading aloud of Italian developmental dyslexics and skilled peers. The stimuli were presented either in mixed blocks (alternating words and nonwords) or in pure blocks. The analyses based on the rate-and-amount model ( Faust et al., 1999 ) indicated that group differences in reaction times between dyslexic and skilled readers (a) were well accounted for in terms of global components and (b) were modulated by context in the case of words but not in the case of nonwords. ANOVAs on z-transformed reaction time data further indicated the influence of stimulus length. Importantly, the frequency effect interacted with context: Controls showed a list context effect for high- and low-frequency words, while dyslexics showed a list context effect only for high-frequency words. The effect of length on reading times remained unaffected by context manipulation. It is proposed that this pattern of results may be accounted for by hypothesizing two separate deficits: An early graphemic impairment affecting performance independently of context and a later inefficiency in activating entries in the orthographic lexicon as a function of context demands.