Results indicated that prevalence rates of retardation were approximately 1 percent when identification was based on individually administered IQ tests (the WAIS or Beta) as contrasted with the 4 percent prevalence rate established by entry-level group IQ testing. Entry-level group testing seriously overestimates the prevalence of retardation in prison. The process by which group tests are administered may contribute more to the increased prevalence rate than the type of test administered. Problems associated with group testing and suggestions for remediation are discussed. Program needs of the retarded are important and must be considered by the prison. It does not help to identify an inmate as retarded if adequate habilitation programs are not available. (Author abstract)