THE STR EVALUATION STUDY WAS INITIATED IN 1974 BY THE NATIONAL HIGHWAY TRAFFIC SAFETY ADMINISTRATION (NHTSA). ITS PURPOSE WAS TO EVALUATE THE EFFECTIVENESS OF ALCOHOL REHABILITATION TREATMENT PROGRAMS USED WITH DEFENDANTS ON DRIVING WHILE UNDER THE INFLUENCE OF ALCOHOL (DWI) CHARGES. FAIRFAX COUNTY WAS ONE OF EIGHT SITES CHOSEN FOR STUDY UNDER A REVISED NHTSA DESIGN. NEARLY 600 PEOPLE WERE DIVIDED ALMOST EQUALLY BETWEEN PREPROBLEM AND PROBLEM DRINKERS FOR THE INVESTIGATION. THEY WERE RANDOMLY ASSIGNED TO VARIOUS ALCOHOL REHABILITATION PROGRAMS INCLUDING POWER MOTIVATION TRAINING (PMT). PREPROBLEM DRINKERS WERE ASSIGNED TO BOTH TREATMENT AND NONTREATMENT CONTROL GROUPS. PARTICIPANTS WERE INTERVIEWED BEFORE TREATMENT AND AGAIN 6 AND 12 MONTHS LATER. INFORMATION ON DEFENDANTS' LIFE SITUATIONS WAS ANALYTICALLY COMBINED INTO A SERIES OF OUTCOME MEASURES INCLUDING THE QUANTITY AND FREQUENCY OF DRINKING, PHYSICAL HEALTH, SOCIAL INTERACTION, AND ECONOMIC AND FAMILY STABILITY. WHILE A LIMITED NUMBER OF POSITIVE SHORT-AND LONG-TERM BENEFITS OF TREATMENT WAS UNCOVERED, A LACK OF CONSISTENT RESULTS WITHIN OR ACROSS DRINKER LEVELS OR BETWEEN THE INTERVIEWS WAS NOTED. REASONS FOR THIS RESULT WERE UNCLEAR ALTHOUGH THEY MAY REFLECT A COMBINATION OF LACK OF VALIDITY/RELIABILITY OF OUTCOME MEASURES. SIGNIFICANT PRETREATMENT DIFFERENCES AMONG PREPROBLEM DRINKERS INDICATING POSSIBLE BIASES IN THE RANDOM ASSIGNMENT PROCEDURES MAY BE TO BLAME AS MIGHT INCONSISTENCIES IN THE DELIVERY OF TREATMENTS. MORE VALIDATION IS SUGGESTED. TABULAR DATA ARE GIVEN AND APPENDIXES OUTLINE THE TEST MEASURES USED. (MJW)