This prospective study was undertaken in order to establish indices of possible neurotoxicity due to the combination of lithium with haloperidol and a combination of lithium with other neuroleptics. Twenty-one subjects who were receiving neuroleptic-lithium combination were studied. Of them, 14 had a primary affective disordermanic phase, and 7 had schizophrenia as judged by Feighner criteria. The subjects were evaluated on clinical, neuropsychological measures and EEG at baseline and 10–14 days after addition of lithium to the neuroleptic regimen. Of 11 subjects receiving the haloperidol-lithium combination, 5 (45.4%) showed abnormal responses on EEG. Of these 5 subjects, 3 showed abnormal photic responses, and 2 showed slowing in posterior alpha activity. These photoparoxysmal and photomyoclonic responses are indicative of cerebral pathophysiology resulting from the interactive effect of the haloperidol-lithium combination. The photic and other associated EEG abnormalities reported may be the earliest indication of neurotoxicity. No significant changes were observed in the EEG in 10 subjects treated with other neuroleptic-lithium combinations. In a statistical comparison the haloperidol-lithium combination had significantly more frequent EEG changes than the combination of lithium with other neuroleptics. This study presents sufficient evidence from reported photic and other associated EEG changes to pursue further investigation of neurotoxicity due to haloperidol-lithium therapy. Such a study should employ a larger number of subjects, random assignment of subjects to treatment and control groups, and blind evaluation of data.