AIMS: This study compared 5-year treatment outcomes of older adults to those of middle-aged and younger adults in a large managed care chemical dependency program. We examined age group differences in individual, treatment and extra-treatment factors, which may influence long-term outcome. DESIGN: Seventy-seven per cent of original study participants completed a telephone interview 5 years after out-patient chemical dependency treatment at Kaiser Permanente. This sample (N = 925) included 65 patients aged 55-77, 296 patients aged 40-54 and 564 patients aged 18-39 (age at baseline). MEASUREMENTS: Measures at follow-up included alcohol and drug use, Addiction Severity Index (ASI), Alcoholics Anonymous Affiliation Scale, social resource and self-reported health questions. Mortality data were obtained from contact with family members of patients as well as automated health plan records. FINDINGS: Older adults were less likely to be drug-dependent at baseline than younger and middle-aged adults, and had longer retention in treatment than younger adults. At 5 years, older adults were less likely than younger adults to have close family or friends who encouraged alcohol or drug use. Fifty-two per cent of older adults reported total abstinence from alcohol and drugs in the previous 30 days versus 40% of younger adults. Older women had higher 30-day abstinence than older men or younger women. Among participants dependent only on alcohol, there were no significant age differences in 30-day abstinence. In logistic regression analysis, age group was not significant. Variables associated with greater age that independently predicted 30-day abstinence in the logistic regression model included longer retention in treatment and having no close family or friends who encouraged alcohol or drug use at 5 years; female gender was also significant. CONCLUSIONS: Results indicate that older adults have favorable long-term outcome following treatment relative to younger adults, but these differences may be accounted for by variables associated with age such as type of substance dependence, treatment retention, social networks and gender. Age differences in these characteristics inform intervention strategies to support long-term recovery of older adults and provide direction for investigation of how age affects outcome.