In a comprehensive rehabilitation center serving the mentally ill 64 dropouts and 42 successful terminees were interviewed monthly for 18 months following termination. The interview schedules elicited information about employment, heterosexual adjustment, living conditions, rent payment, social preferences, intimate friendships, and rehospitalization episodes. Analysis revealed that dropouts and successful terminees typically differed significantly on these variables for at least part of the follow-up period, but that by the end of the follow-up period there were no significant differences between the two groups on any of the variables, whether or not rehospitalization was controlled for. This hindsight examination of dropouts in comparison with successful terminees shows the concept of "dropping out" to be a time-limited concept.