Opiate withdrawal symptom ratings by trained clinicians were compared to self-reports of withdrawal discomfort during an outpatient, blinded, randomized clinical trial of clonidine detoxification for methadone maintained subjects. The randomized comparison group of subjects were detoxified using a slow tapering of methadone over 1 month. For all 39 subjects in this study the observer and self-ratings were substantially correlated (γ=. 75), but moderate levels of disagreement also occurred and the observer ratings were often lower. The difference in actual withdrawal scores and amount of shared variance between the observer and self-ratings were used as indices of disagreement for each individual subject. We found that detoxification failures had significantly less interrater agreement than the successes and that subjects who reported more distress than the observers noted were more likely to fail at detoxification. Subjects being detoxified using clonidine were more likely to fail at detoxification. Subjects being detoxified using clonidine were more likely to demonstrate this association between failure and disagreement on self versus observed withdrawal than were the methadone group. We concluded