Many branches of medicine rely heavily on lab tests to monitor client treatment response and use this information to modify their treatment. By contrast, those who offer psychological interventions seldom rely on formal assessments (lab tests) to monitor their clients' response to treatment. Data are presented that demonstrate that clinicians rarely accurately predict who will not benefit from psychotherapy. This finding is contrasted with the use of a questionnaire (lab test data) and decision rules on the basis of a client's expected progress. Results have indicated that formal methods of monitoring were able to identify 100% of the patients whose condition had deteriorated at termination, and 85% by the time they had attended three treatment sessions. Practitioners are encouraged to consider formal methods of identifying the deteriorating client. © 2004 Wiley Periodicals, Inc. J Clin Psychol/In Session.