An experiment was designed to test an instrument that could assist writers and editors in selecting strategies for improving the comprehensibility of written instructions. In addition, the study examined how successful professional writers are in improving passages selected for difficulty. The subjects, 150 experienced writers, were given an original passage to read and rate using the instrument. They then completed a multiple choice test for the passage. They were to revise the passage and rate their version. More than 500 students were then given either the original or the revised version of the passage and were tested on their comprehension of it. The findings tended to support traditional notions of the factors that contribute to effective written instruction, such as the importance of content and of student aptitude. The most suprising result of the study was the failure of the writers as a group to improve the comprehensibility or memorability of the two passages, even though the passages had been chosen because they seemed difficult to understand. This was true even though the writers were generally professional developers of instructional materials, even when they had in hand a set of guidelines for judging the comprehensibility of the passages, and even when they knew they were participating in an experiment related to the comprehensibility of passages. (Materials used in the study are appended.) (EL)