Forty-three high school students (14 poor, 13 average, and 16 good readers) were given three different measures of memory capacity to determine if they differed in working memory, traditional memory span, and chunking capacity. Several experimental tasks followed to determine if differences in the memory factors were related to specific components of reading. These tasks included lexical decision, letter matching, reading rate, synonym match, speeded comprehension, and word recoding. Comprehension, IQ, spelling, mathematics word problems, and vocabulary scores were collected from school records. The three memory tasks, lexical decision, and synonym match discriminated significantly among the groups, with general ability controlled. Regression analyses suggested each memory variable was explained by a different constellation of the other reading factors. Findings suggest that the relationship between working memory and reading comprehension is not isolable from the type and amount of information encoded during its measurement. (One table of data is included.) (Author/RS)