This investigative report analyzes 1978-83 enrollment data from two national studies, selected state studies (California, Florida, Massachusetts, and New Jersey) and local school district studies (New York City, Los Angeles, Chicago) in order to determine the prevalence of exceptional bilingual students, examine general trends, and report recommendations for future data collection. Among the variables cited as affecting the accuracy of prevalence rates for culturally and linguistically diverse exceptional (CLDE) students are changing CLDE definitions, misdiagnosis, sampling error, the role of the individual school in referral and placements, and reluctance to stigmatize CLDE students as handicapped. Review of the national studies leads to such findings as that district size and racial composition correlate to disproportionate special education enrollment, and that districts with the highest disproportion levels tend to have the smallest proportion of students in bilingual programs. Among conclusions of the analysis of the state studies is that where the Hispanic population makes up more than 1% or more of the state's total population there is a trend toward underenrollment of Hispanics in gifted and talented programs and over-enrollment in classes for the mentally retarded. Analysis of the local district studies also revealed a disproportionate representation of limited-English-speaking or non-English speaking pupils in special education programs. One of 11 recommendations is that the school districts should collect and analyze information regarding the number of referrals to special education by language proficiency level within each ethnic group. (JW)