Although neuroleptics are the major treatment for schizophrenia, there are no effective measures to determine the appropriate or necessary length of neuroleptic maintenance. To test the ability of a psychostimulant challenge to predict relapse following neuroleptic withdrawal, the authors administered methylphenidate and placebo infusions to 11 stable schizophrenic outpatients who had been on a neuroleptic maintenance regimen for at least 6 months. Patients withdrawn from neuroleptics were followed until relapse. All three patients with a positive response to methylphenidate challenge relapsed in 1 to 7 weeks; one of seven negative responders relapsed at week 21. Differences in relapse rates (p = .033) and survival time (p = .005) between negative and positive responders were significant.