Summary.This study investigates whether good and poor readers use different strategies when making use of the linguistic context, and in particular whether good readers make better use of the succeeding context by using a better strategy or whether they do so simply because of their superior knowledge. A cloze test was specially constructed to control for the readers' knowledge and, after successive refinements, was administered to 121 7‐ and 8‐year‐old children in their first year at junior school. The children were better at using the preceding context than the succeeding context (P<0.001), and the better the reader the better was his use of the succeeding context (P<0.01). As the knowledge of the readers had been controlled this must have been a consequence of some difference in the strategies used by the better and poorer readers. It was not, however, because the good readers made better use of an overt self‐corrective strategy, as self‐corrections were not found to be related to reading ability. No conclusion could therefore be drawn as to whether the good readers used a more efficient strategy, or whether they used the same strategy more skilfully. It was also argued that, as poor readers are lessableto use the context, it might well be a mistake to conclude that they arereluctantto do so when there is graphic information available simply on the evidence that they make less