The "yes/no" method of vocabulary assessment requires students to indicate words they know from among a list of words and nonwords. Preliminary evidence gained from a study involving fifth grade students indicates that the method is superior in many ways to the multiple choice method of assessment. Analysis of "false alarms," cases in which children say they know the meanings of nonwords, reveals that good readers aggressively apply morphological rules to hypothecate meanings for unfamiliar terms, whereas poor readers engage in phonemic experimentation with unfamiliar items to transform them into common words. A review of the literature shows that vocabulary difficulty is a factor in text comprehension, but that it is not as important as studies of readability have suggested. (FL)