Tardive dyskinesia (TD) affects between 10 and 50 percent of all patients on long-term antipsychotic therapy, depending on the population studied. Various risk factors for TD have been reported; a correlation between TD and substance abuse has been suggested in some reports and not found in others. This study analyzes the association of substance abuse with the incidence of tardive dyskinesia in a schizophrenic population. All patients at the West Side Veterans Affairs Medical Center are evaluated prior to the initiation of neuroleptic therapy with the Dyskinesia identification System: Condensed User Scale (DISCUS); those with a diagnosis of schizophrenia, schizoaffective disorder, or schizophreniform disorder during the years 1986 through 1993 were included in this analysis. History of substance abuse was considered positive if there was clinician report or diagnosis of substance abuse. These data were collected and analyzed using ANOVA. In a sample of 1,027 subjects (97% male), 83.2 percent had a neuroleptic exposure of 10 or more years, and slightly more than half (50.8 percent) had a positive history of substance abuse. Using research diagnostic criteria, 28.9 percent of the sample had tardive dyskinesia. Analysis of variance showed history of substance abuse (p < .000) and years on neuroleptics (p < .000) to be strongly correlated to a diagnosis of TD. Age was less strongly correlated to the DISCUS score (p < .01), and there was no association of TD with diagnosis (p = .237). This study therefore demonstrates a robust correlation between TD and substance abuse. A mechanism of action involving N-methyl-D-aspartate (NMDA)-mediated excitotoxicity is proposed.