Whether grouping children into special classes on the basis of IQ scores is educationally defensible was investigated using 492 educable mentally retarded (EMR) students, 12-15 years of age, during two field tests (1971-73). The curriculum used was the Me and My Environment biological sciences program. The following three test instruments were developed and administered: the Cognitive Development/Problem Solving (CD/PS) Test of Developmental Level, teacher ratings of students, and student performance measures for program units 1, 2, and 4. Analysis of data was achieved through 28 regression analyses involving 17 variables such as the student performance measure, the four CD/PS subtests, six teacher ratings, demographic variables, and test class data. Results showed that it is possible to measure levels of cognitive development with the CD/PS Test without the measure of IQ. Factors of IQ, age, sex, and ethnic background accounted for little or no explained variance over seven replications. Implications included exploration of strategies such as developmental tests for assessing performance abilities of EMR children. (An appendix which comprises more than half the document contains tables of statistical data and test examples.) (MC)