A study was conducted to examine the impact of various role induction modalities in dealing with the problem of early dropout among substance abusing patients. Emphasis was placed on tracing the interdependency between patient misperceptions of treatment, termed level of discrepancy in treatment expectations (Level of DTE) and continuance. Results indicated that patients who were more favorably disposed to the program (low Level of DTE) were more likely to remain in care than those who were not (high Level of DTE) only if role induction was provided; in the absence of the pretreatment intervention, differences between high and low discrepant patients were virtually nonexistent. These data suggest that the real utility of role induction lies in its ability to confirm the expectations of low discrepant patients (in contrast to improving patient knowledge of therapy) while at the same time providing them with a mechanism to learn coping skills necessary for adjusting to a new program. For those classified as high discrepant patients, role induction alone was not sufficient to prevent early termination; the majority did not stay past four sessions. Implications for this group are that greater efforts should be made in responding to their concrete problems, actively engaging“significant other(s)”or providing them with an external support system in the form of a“buddy”at the point of entry, in order to forestall premature termination.