Verbal responses to several tests of originality were subjectively rated for originality. The ratings were compared with the statistical frequency of these responses in two samples of test takers, 150 elementary school children and 60 college students. The elementary school children were administered three measures from the Wallach and Kogar test battery: Uses, Instances, and Similarity. The children were encouraged to think of a response nobody else in their class would think of. The college students' tests were the Alternate Uses test of Christensen, Guilford, Merrifield, and Wilson and the Consequences test of Christensen, Merrifield, and Guilford. Responses were classified in four categories: unique, used twice, used three times, and used four or more times. Responses were also rated on a five-point scale of originality. Interjudge reliability of the ratings ranged from .62 to .92 for various samples; but the correlation between originality ratings and infrequency of usage ranged from .03 to .16. It is suggested that response frequency scores are not a valid substitute for originality scores for tests like these. (CTM)