Longer interval (M = 50.6, SD = 13.1 days) test-retest reliability of the l
ifetime items on the Addiction Severity Index (ASI), a semistructured inter
view, was evaluated in 108 alcohol and/or cocaine dependent patients. They
were administered the ASI at admission to an intensive outpatient rehabilit
ation treatment program and again after completion of this intervention and
randomization into an aftercare study. Results demonstrated good to excell
ent reliability for participants' reports for most lifetime items in the me
dical, employment, drug, alcohol, and legal problem areas. Two of the ASI a
reas, family/social and psychiatric, had numerous items that did not achiev
e acceptable levels of reliability. Within these two problem areas, the mor
e subjective and less operationally defined constructs had the poorest reli
ability. This study, in general, supports the longer interval test-retest r
eliability of the ASI lifetime items as well as the notion that alcohol and
cocaine dependent patients under certain conditions can and do reliably re
port personal information.