An earlier study was extended and replicated to examine the feasibility of generating multiple-choice test questions by transforming sentences from prose instructional material. In the first study, a computer-based algorithm was used to analyze prose subject matter and to identify high-information words. Sentences containing selected words were then transformed into multiple-choice items by four writers who generated foils or question alternatives informally and by an algorithmic method. These items were then organized into tests and administered to 24 college students before and after they had studied the instructional materials. In this replication, the tests were administered to 249 high school students, and results were combined with those obtained earlier. This provided stable estimates of item difficulty. Results supported those obtained earlier. Thus, it appears that this item-writing technique is feasible and that algorithmic methods of generating foils produce items of reasonably good quality. (The prose passage used in the study and examples of test items are appended). (Author/CTM)