This study developed procedures for teachers to use in ranking the reading difficulty level of student-authored materials in the primary grades. Two procedures were developed: the word procedure for assessment of reading difficulty levels of vocabulary, and the two-four procedure for assessing reading difficulty level of sentence structure. Because the word procedure was found to be more reliable and practical than the two-four procedure, it was subjected to further evaluation of validity for predicting the reading difficulty level of stories written by children for other children to read. Four student-authored stories, varying in length and in word-procedure rankings of difficulty, were chosen from collections of children's writing. Twenty second-grade children read the stories, and their comprehension was assessed by an analysis of their spontaneous retelling of the stories. Readers were asked to rank the stories from easiest to most difficult. The rankings agreed with word-procedure rankings of reading difficulty and with mean retelling scores for each story. It was concluded that the word procedure could be useful in predicting reading difficulty level if stories being ranked were approximately equal in length. (Author/LL)