Studies in nonhuman primates with experimental parkinsonism have shown that intrastriatal cografts of autologous adrenal medulla and peripheral nerve yield greater behavioral improvement and graft survival than do adrenal medulla grafts alone. To test these observations, five patients with advanced Parkinson's disease were selected to receive unilateral intrastriatal adrenal medulla-intercostal nerve cografts. They were evaluated using the Core Assessment Program for Intracerebral Transplantation (CAPIT) protocol. Three of these patients also underwent quantitative motor testing for the measurement of upper limb bradykinesia (movement time; MT). Following right flank adrenalectomy, cografts consisting of small fragments of adrenal medullary tissue and minced intercostal nerve were stereotaxically implanted into three targets in the right striatum using computerized tomography guidance. Surgery was uneventful and postoperative magnetic resonance imaging revealed accurate placement of the grafts. No morbidity was encountered. Results of 24 months of clinical and quantitative motor assessments postoperatively are reported. Total UPDRS motor scores in the 'off' state improved from a mean preoperative score of 39.5 to 32.1 at 3, 29.7 at 6, 27.6 at 9, 28.5 at 12, 31.4 at 18, and 26.5 at 24 months after surgery. Total timed motor test scores during the 'off' state improved 17.9% at 6, 23.3% at 9, 18.2% at 12, 38.2% at 18, and 34.9% at 24 months postoperatively compared to baseline. Movement time showed statistically significant improvement (repeated measures ANOVA, P < 0.05) in the left arm (contralateral to surgery) in all three patients tested. These results indicate that stereotaxic intrastriatal implantation of autologous adrenal medulla- peripheral nerve cografts can be performed safely and clinical improvement from this procedure is sustained for a period of 24 months. The clinical improvement was paralleled by improvement in objective, quantitative motor testing.