Intrastriatal grafting of dopamine-rich embryonic ventral mesencephalon (VM) is a potential therapeutic treatment for Parkinson's disease. However, it has been suggested that the efficacy of this procedure might be improved by enhancing the survival and/or degree of neurite outgrowth by the grafted VM, since these parameters are currently suboptimal. In the present study, we tested the ability of astrocytes retrovirally transduced to produce recombinant brain-derived neurotrophic factor (BDNF) to enhance the survival and/or function of embryonic VM in the unilateral 6-hydroxydopamine (6-OHDA) lesioned rat, a well-characterized rodent model of Parkinson's disease. In culture, primary astrocytes derived from Postnatal Day 0 (P0) rat striatum and transduced with the BDNF vector increased the survival of Embryonic Day 15 (E15) dopaminergic VM neurons by approximately threefold and reduced the loss of dopaminergic neurons following 6-OHDA treatment by approximately 20%. The cultured astrocytes were then mixed 1:1 with freshly dissociated E15 VM and co-grafted into the dopamine-denervated striatum. Unexpectedly, the control nontransduced astrocytes reduced the survival of dopaminergic neurons by 60% and restricted the pattern of neurite outgrowth by the co-grafted VM, compared to grafts of VM alone at 7 weeks postgrafting. These effects were paralleled by an attenuated rate and degree of behavioral recovery. The detrimental effects of the control astrocytes were partially reversed when the astrocytes were transduced to express BDNF, although dopaminergic neuron survival was still reduced by 30% compared to that within VM-only grafts. To begin to assess whether the detrimental effects of the astrocytes were related to the maturational state of the cultured astrocytes, astrocytes were obtained from E18 striatum and maintained in short-term culture (9 days vs several weeks for P0 cultures) prior to co-grafting with VM. Interestingly, the younger astrocytes did not reduce graft survival and allowed for better graft integration. These results suggest that primary astrocytes maintained in long-term culture are detrimental to embryonic neural grafts, an effect that is not completely overcome by expression of recombinant BDNF, and that astrocyte age may be an important consideration in the use of these cells as CNS gene delivery vehicles.