It has been suggested that levodopa (L-dopa), a dopamine precursor used to treat Parkinson's disease, may be toxic to grafted fetal neuroblasts; if so, the use of the monoamine oxidase B inhibitor selegiline might prevent such toxicity. We randomly assigned 30 unilaterally 6-hydroxydopamine-lesioned male Sprague-Dawley rats, whose lesions were verified with low-dose apomorphine-induced rotations, to one of five treatment groups: (i) L-dopa methyl ester (125 mg/kg/day) with benserazide (a peripheral decarboxylase inhibitor; 25 mg/kg/day), (ii) L-dopa methyl ester with benserazide and selegiline (L-deprenyl; 0.5 mg/kg/day), (iii) selegiline only, (iv) and (v) vehicle (ascorbate in normal saline) only. After 2 weeks of twice-daily ip injections, the rats received fetal ventral mesencephalic grafts into the lesioned striatum; one vehicle group received sham grafts. Drug therapy was contin ued for 21/2 months post grafting. At 1 month after grafting, and every 2 weeks thereafter, the rats were tested using low-dose apomorphine-induced rotation. A 70% decrease in rotations among all grafted groups, relative to the shams, was found. No statistical differences among groups receiving various drug therapies were seen in behavior or in counts or dimensions of tyrosine hydroxylase-positive cells. We therefore conclude that, in the unilaterally lesioned rat model of Parkinson's disease, there is no adverse effect of L-dopa nor any significant effect of selegiline, either alone or coadministered with L-dopa, on behavioral recovery induced by fetal ventral mesencephalic grafts. (C) 1994 Academic Press, Inc.