College students read a text either with or without inserted questions. Question groups performed better, relative to controls, on post-test items that repeated inserted questions, and on new post-test items from the same categories as the inserted questions. A selective attention interpretation of the effect of inserted questions was made. (Author/RD)