Although vocabulary tests have not been well suited to subdivisions that yield scores for different attributes, they have often been used in this manner. The question of whether such subscores can be justified from a measurement standpoint was studied. Items of a vocabulary test were classifed into three categories; the words were presented in isolation, in phrase context, or in passage context. Subscores based on these three kinds of tasks were obtained from samples of approximately 3,000 students in each of grades 3 through 9. The subscores were then correlated. After correction for attenuation, the intercorrelations clustered sufficiently close to 1.00 to suggest that the separate subscores would not likely serve productive purposes. The two kinds of questions most frequently found in standardized vocabulary tests are isolated words and words in phrase context. The findings of this study indicated that these two item types offer alternative ways of measuring the same thing. It does not seem to matter whether a vocabulary test presents vocabulary words in isolation, in phrase context, or in a combination of the two. (Author/RD)