The test-retest reliability of alcohol abusers' self-reports of their daily drinking and daily drinking, alcohol-related incarcerations and their drinking problem history were highly reliable (r = +0.79 to +0.98). Limits on the generalizability of these finclings are discussed. ing dispositions for the 360-day period preceding admission to treatment. Results indicated that. over a 6-week test-retest interval, outpatient male alcohol abusers' self-reports of their daily drinking, alcohol-related incarcerations and their drinking problem history were highly reliable (r = +0.79 to +0.98). Limits on the generalizability of these finclings are discussed.