Despite the proven efficacy of neuroleptic drugs in the acute and maintenance pharmacotherapy of schizophrenia, practical methods for identifying patients who require neuroleptic treatment to prevent relapse are lacking. This study evaluated the use of a methylphenidate challenge test to predict the outcome in 34 stable outpatients with schizophrenia receiving neuroleptic treatment. Patients received two infusions, one of methylphenidate and one of placebo, in randomized order one week apart while receiving neuroleptic treatment and again three weeks after drug withdrawal. Behavioral, cardiovascular, and neurologic responses were evaluated before and after infusion under double-blind conditions. Patients were then followed up without medication for 52 weeks or until symptom recurrence. The results indicate that specific measures, including behavioral response to methylphenidate, presence of tardive dyskinesia, and, under specific pharmacologic conditions, tardive dyskinesia, blink-rate, and pulse-rate responses to methylphenidate, are associated with time and propensity to relapse following neuroleptic withdrawal. These measures may be potentially useful in the identification of candidates for neuroleptic withdrawal and/or dosage-reduction treatment strategies.