In order to investigate differences between the comprehension of high or low imagery passages of long or short length, a study analyzed the reading comprehension of fifth grade pupils (all reading at grade level). Thirty subjects read 16 passages each: four high imagery/short; four high imagery/long; four low imagery/short; and four low imagery/long. Imagery of a passage was determined by matching the imagery level of words in each passage with two predetermined word lists with imagery ratings. Comprehension questions consisted of three factual and three inferential questions written and field tested for each passage. Pupils, tested individually, read each passage silently. A verbatim record of oral answers to questions was recorded for later scoring. Results showed that total comprehension of high imagery passages was significantly greater than low imagery passages regardless of length. Factual questions were generally answered significantly better than inferential questions. The study suggests that imagery as a mediator appears to be one of the determinants of comprehension. (Four tables of data are included; 24 references are attached.) (SR)