SummaryAlthough aftercare following a specified treatment event is gaining popularity, its effect on long term outcome is uncertain. Cross‐lagged panel analysis (correlation) and path analysis (standardised partial regression) are two methods to examine the relationship. These methods were used to describe the aftercare—outcome effect at two points in time, one and two years following inpatient treatment of 37 Caucasian male alcoholics. Social stability at intake and behavioural adjustment while hospitalized were added to the path model to help rule out certain confounds of the aftercare variable. Aftercare influenced the outcome sys