INDIVIDUALIZED GAS HAS BEEN PROPOSED AS A PROCEDURE TO ALLOW FOR INDIVIDUALIZED CLIENT GOALS WHILE PROVIDING DATA FOR USE IN EVALUATING AN ENTIRE PROGRAM. THE BASIC APPROACH OF GAS THUS CALLS FOR DEFINING A UNIQUE SET OF GOALS FOR EACH CLIENT AND/OR PROGRAM RATHER THAN DEFINING MORE GENERAL PROGRAM GOALS TO BE MEASURED ACROSS ALL CLIENTS AND/OR PROGRAMS. WITHIN THE GAS FRAMEWORK, THE CRITERION OF PROGRAM SUCCESS BECOMES THE EXTENT TO WHICH INDIVIDUAL GOALS ARE ACHIEVED. THREE STEPS ARE BASIC IN THE GAS METHODOLOGY: (1) ASSIGN CLIENTS TO A DESIGNATED INTERVENTION; (2) SELECT GOALS AND CREATE GAS SCALES; AND (3) CONDUCT A FOLLOWUP ASSESSMENT OF EACH CLIENT'S STANDING ON GOALS AFTER TERMINATION FROM A PROGRAM. ALTHOUGH GAS CAN BE AN EFFECTIVE THERAPEUTIC TOOL, MORE RESEARCH IS NEEDED TO DETERMINE MORE PRECISELY CONDITIONS UNDER WHICH GAS HAS ITS GREATEST THERAPEUTIC IMPACT. IN EVALUATING THE THERAPEUTIC EFFECTIVENESS OF GAS, IT IS ESSENTIAL THAT OUTCOME CRITERIA THAT ARE INDEPENDENT OF THE GOAL SETTING PROCESS ITSELF BE USED. GAS IS DEFICIENT ON A NUMBER OF PSYCHOMETRIC DIMENSIONS, INCLUDING THE CONTENT OF CLIENT GOALS, RELIABILITY, AND VALIDITY. LIMITATIONS OF GAS WITH RESPECT TO RESEARCH DESIGN PRIMARILY RELATE TO CONTROLLING FOR SELECTION BIASES USING GAS VERSUS UNIFORM MEASURING INSTRUMENTS. BECAUSE POLICY IMPLICATIONS OF GAS SCORES ARE NOT ALWAYS CLEAR, GAS IS NOT A RELIABLE EVALUATION TOOL. ONE SOLUTION TO THE APPARENT CONFLICT BETWEEN THERAPY AND EVALUATION IS TO EMPLOY BOTH INDIVIDUALIZED GAS AND UNIFORM ASSESSMENT DEVICES. REFERENCES ARE INCLUDED. (DEP)