The subjects were 85 persons who had been convicted of DWI offenses and were required by the courts to attend a 10-hour program of instruction. The program was intended to increase subjects' knowledge about alcohol and its effects. Subjects were evaluated for alcoholism, and those identified as alcoholics were referred for treatment. The subjects' knowledge level was determined by a post-test and knowledge inventory test. Alcoholism was determined by three separate diagnostic tests, and records from treatment facilities for those alcoholics referred were screened. Driving records for each subject were reviewed to determine subsequent arrests for DWI within 12 months after completion of the program. A 2x2x2 factorial research design was used. No significant relationship between knowledge level and subsequent arrests and between referral for treatment and subsequent arrests was found. However, a significant relationship was found between the presence of alcoholism and subsequent arrests. Additional information on arrest rates subsequent to treatment and other factors is provided. Recommendations are made for more research concerning the effectiveness of DWI schools. Study instruments and approximately 25 references are appended. (Author abstract modified)