Two studies examined the relationship between word knowledge and reading comprehension. Subjects were college undergraduates with high and low verbal abilities as indicated by a standardized verbal aptitude test. The first study involved a multiple choice vocabulary test from which words that both groups defined correctly were selected. The subjects were then asked to use the words in sentences, define their meanings, and make semantic comparisons among them. The results indicated that low verbal subjects' knowledge of word meaning was less complete and more bound to a specific context than that of high verbal subjects. The second study investigated this possibility with a combined protocol and reaction time methodology. Subjects read two paragraphs--one using a word in a familiar context, the other in an unfamiliar context--and were asked to define the word. The second study indicated that the completeness of knowledge about the meaning of a word that is familiar seems to affect both sentence reading time and paragraph recall (although the former more so for the high verbals and the latter more so for the low verbals). The results suggest that, although it may be a relatively straightforward task to teach a connection between an "unknown" word and a "known" word, it is a much more complex matter to provide and promote the decontextualized knowledge that will facilitate the comprehension of those words in an unfamiliar text. (HTH)