Two paper and pencil group tests were assessed for their ability to classify subjects as concrete or formal operational. The two group tests, Sheehan's (1970) version of Longeot's three subtests and set of three puzzles, were administered to 607 students from ninth- and tenth-grade science and mathematics classes. A subsample of 69 students was involved in clinical interviews on a battery of three Inhelder tasks. Statistical analysis of test results were used to assess the correlations among subtests and test batteries and to provide a measure of the validity of the group tests. It was concluded that the tests used did provide a measure of concrete versus formal operations, but that both tests appeared to inflate the number of formal subjects. (Author/GK)