METHOD 1, CALLED 'VISUALLY IDENTIFYING NEEDS,' IDENTIFIES FACTORS REVEALED IN SOCIAL STATISTICS CONSIDERED TO CORRELATE WITH NEED FOR MENTAL HEALTH SERVICES. MAPS ARE DEVELOPED FOR THESE VARIABLES REPRESENTING DIFFERENT LEVELS (PERCENTAGES OR RATES) OF EACH FACTOR IN EACH GEOGRAPHIC AREA. WHEN THESE MAPS ARE TRANSFERRED ONTO TRANSPARENT PLASTIC SHEETS AND SUPERIMPOSED ON ONE ANOTHER, THEY SHOW GEOGRAPHIC CONCENTRATIONS OF APPARENT SERVICE NEEDS. UTILIZATION RATES CAN THEN BE RELATED TO AREAS OF HIGH RISK TO DETERMINE WHETHER A PROGRAM IS REACHING ITS TARGET POPULATION. LACK OF CONCENSUS ON SOCIOECONOMIC VARIABLES RELATED TO MENTAL ILLNESS AND DECIDING WHICH VARIABLES MAY MOST ACCURATELY PREDICT MENTAL HEALTH NEEDS ARE CONSIDERED THE MAIN PROBLEMS WITH THIS APPROACH. METHOD 2, CALLED 'IDENTIFYING THE RELATIVE AMOUNT OF NEED IN EACH CENSUS TRACT,' ASSUMES THAT SOCIAL STATISTICS ARE ACCURATE PREDICTORS OF LEVEL-OF-NEED, AND ASSIGNS TO EACH CENSUS TRACT A PERCENT OF THE TOTAL NEEDS OF THE GEOGRAPHIC AREA. THIS METHOD TAKES THE LEVEL OF NEED IDENTIFIED BY THE PREVIOUS METHOD FOR EACH CENSUS TRACT AND MULTIPLIES THE LEVEL OF NEED BY THE NUMBER OF PERSONS IN THE TRACT. THIS METHOD IS CONSIDERED TO MEET WITH THE SAME PROBLEMS AS THE PREVIOUS METHOD. METHOD 3, 'IDENTIFYING NEEDS IN ABSOLUTE NUMBERS,' ASCERTAINS THE ABSOLUTE NUMBER OF PERSONS SUFFERING FROM A SPECIFIC DISORDER. IN USING THIS METHOD, AS MUCH DATA AS POSSIBLE MUST BE GATHERED, AND THEN AN EDUCATED GUESS MUST BE MADE. FOUR EXAMPLES OF THE USE OF THIS METHOD, EMPLOYING SOCIAL STATISTICS DIFFERENTLY, ARE PROVIDED. REFERENCES ARE INCLUDED. (RCB)