Haloperidol administration for 2 weeks results in significant reduction in the concentration of GSH in the CSF. Concomitantly, the levels of lipid peroxidation products increased as evidenced by increased malondialdehyde levels. The malondialdehyde levels in the CSF prior to haloparidol administration were not significantly higher than that seen in CSF from normal controls (data not shown) suggesting that increased oxidative stress did not exist in these patients prior to haloperidol administration. All the patients included in the present study were drug naive and hence the changes observed in the glutathione and malon6ialdehyde levels in the CSF were indeed mediated by haloperidol administration. The only other medication that was administered namely, anticholinergic drug, trihexyphenidyl is not known to cause any oxidative stress. The present study thus demonsuates that haloperidol administration results in significant oxidative stress. The generation of the oxidative stress is probably due to the increased turnover of dopemine caused by typical neuroleptics. Increased dopamine turnover is also observed in Parkinson's disease and the combination therapy consisting of antioxidant vitamin E and monoamine oxidase inhibitor, deprenyl has been shown to offer limited protection against the progression of the disease (Parkinson Disease Study Group 1989). In the present study, all the 15 patients exhibited extrapyramidal symptoms although the time of onset, the duration and the severity of the side effects differed between patients. On the presumption that the oxidative stress generated by haloperidel may cause extrapyramidal symptoms, the present study in humans taken together with the evidence provided in our earlier studies on rats (Shivakumar and Ravindranath 1992,1993) may justify experimental coadministration of antioxidanls (e.g., vitamin E) with typical neuroleptics like haloperidol to prevent the acute side effects.