The Yes/No vocabulary test has some methodological advantages compared with the commonly used multiple-choice vocabulary test. To clarify the nature of both tests, we conducted two surveys. Participants ranged from high school students to adults in their 60s. We examined the relationships between these two vocabulary tests and reading activity, which is a major source of vocabulary learning. We found an intermediary relationship: familiarity with reading directly predicts known rates on the Yes/No test, and the known rate predicts correct answer rates on the multiple-choice test. Based on this result, we concluded that the Yes/No test reflects a “rough understanding of the meaning of words,” which is the initial vocabulary learning effect of reading. Thus, the effect of reading activities can be detected more sensitively by the Yes/No test. The results also suggest that the multiple-choice test reflects a “detailed understanding of the meaning of words” that precedes the rough understanding of meaning.