THE TECHNIQUE FOR DEVELOPING THE MODEL RELIED ON A PREVIOUSLY CONDUCTED SURVEY. THE PROBLEMS INHERENT IN FINDING A SUITABLE SURVEY CONTAINING SUFFICIENT QUANTITIES OF RELIABLY COLLECTED DATA ARE DISCUSSED. THE TYPES OF DRUG ABUSE INFORMATION GENERALLY YIELDED BY COMMUNITY SURVEYS ARE EXAMINED. SAMPLING ERROR, THE SAMPLING ERROR OF THE BASE CENSUS INFORMATION, AND THE GEOGRAPHICAL UNIQUENESS OF AN AREA ALL COMBINE TO MAKE SURVEYS FOR AREAS WITH POPULATION BASES SMALLER THAN SEVERAL THOUSAND HIGHLY SUSPECT. WITH THIS IN MIND, THE STUDY SELECTED A NATIONWIDE SURVEY CONDUCTED BY THE SOCIAL RESEARCH GROUP AT GEORGE WASHINGTON UNIVERSITY (WASHINGTON, D.C.) UNDER A CONTRACT FROM THE NATIONAL INSTITUTE OF DRUG ABUSE AS THE BEST SOURCE OF SURVEY INFORMATION. THE SURVEY, COMPLETED IN 1974 AND PUBLISHED IN 1975, WAS TECHNICALLY APPROPRIATE AND INFORMATION WAS AVAILABLE BY FOUR REGIONS OF THE UNITED STATES, OF WHICH THE WESTERN REGION INCLUDING OREGON, WAS USED. A NONMATHEMATICAL DISCUSSION OF THE DEVELOPMENT OF THE MODEL IS GIVEN. THE MODEL USES THE SURVEY'S ABUSE RATES BY SUBSTANCE (ALCOHOL, OPIATES, DEPRESSANTS, STIMULANTS, AND OTHER DRUGS) TO DETERMINE POTENTIAL CLIENTELE FOR SUBSTANCE ABUSE PROGRAMS IN OREGON. AGE DISTRIBUTION AND DEMOGRAPHIC VARIABLES ARE ALSO CONSIDERED. THE ESTIMATES OBTAINED THROUGH THE MODEL ARE THEN COMPARED WITH THE ACTUAL EXPERIENCES OF MAJOR OREGON ALCOHOL AND DRUG ABUSE PROGRAMS. THE PRELIMINARY VALIDATION IS HIGHLY SATISFACTORY. IT IS SUGGESTED THAT, AFTER FURTHER REFINEMENT, THIS MODEL CAN BE USED TO ESTIMATE POTENTIAL CLIENTELE FOR SUBSTANCE ABUSE PROGRAMS BASED ON STANDARD CENSUS DATA. TABLES PRESENT FINDINGS, AND REFERENCES ARE APPENDED. (GLR)