A study measured the effects that direct instruction in narrative text structure using a story map (called "macrostruction") had on the comprehension of average and above average sixth-grade readers. A 2 x 2 x 2 repeated measures design was used with time, text, and treatment as the independent variables. The number of story grammar idea units recalled served as the dependent variable for free- and prompted-recall, with a 4-point rating scale used to categorize the type and frequency of responses given to short answer interpretive and applicative questions. Twenty-one subjects received a 4-week treatment in identifying and using the underlying structure of a text for improving reading comprehension, while a control group continued with routine classroom instruction. Contrary to similar studies conducted, the instructional treatment did have a significant effect upon the number of idea units recalled and the type of comprehension of better readers. Hypotheses concerning the amount and type of comprehension affected were supported. Possible explanation for these results are discussed and implications for teachers and researchers stated. (Author)