An investigation was made of the effects of the removal of certain grammatical categories from standardized reading comprehension test paragraphs on the reading comprehension of 180 third, fourth, and fifth graders enrolled in a rural Georgia school system. Each grade level contained a high and low reading ability group, based on standardized reading test scores. The materials used were paragraphs and the accompanying multiple-choice items from various standardized reading tests appropriate for these grade levels. The paragraphs were altered by deletion of the following grammatical elements: (1) all nouns, (2) all adjectives, (3) all verbs, (4) all function words, (5) all words, and (6) no words. The deletion of nouns had a significant effect at all three grade levels, but the deletion of adjectives had a significant effect only for fifth grade. It was also found that the subjects could answer the multiple-choice items without the aid of the accompanying reading paragraph at a level greater than chance. Therefore, it was concluded that the reading comprehension test was measuring other variables such as past learning and word association in addition to ability to take cues directly from the reading paragraph. Tables and references are included. (Author/CM)