Studies in the literature that attempt to relate neuroleptic plasma levels to the development of tardive dyskinesia (TD) report inconsistent findings. As part of an open, long-term study, 60 schizophrenic and schizoaffective patients were started gradually on a b.i.d. schedule of the atypical antipsychotic drug clozapine. Blood samples were drawn weekly for 6 weeks and analyzed for a variety of constituents including clozapine plasma levels. Patients with higher levels of TD were found to have significantly higher levels of plasma clozapine and a higher ratio of plasma/dose than those with lower levels of TD. Our data suggests that schizophrenics with TD may have different pharmacokinetics, drug metabolism, and elimination processes than those without TD. Higher typical plasma neuroleptic levels may increase susceptibility to TD development. A second hypothesis implies that it is not the higher mean plasma level of a neuroleptic that is associated with TD but the greater fluctuations of plasma levels over time (i.e., a higher variance). This hypothesis is discussed in the context of our data.