In an attempt to improve the survival of implanted dopamine cells, we have readdressed the optimal embryonic donor age for dopamine grafts. In a rat model of Parkinson's disease, animals with unilateral 6-hydroxydopamine lesions of the median forebrain bundle received dopamine-rich ventral mesencephalic grafts derived from embryos of crown to rump length 4, 6, 9, or 10.5 mm (estimated embryonic age (E) 11, E12, E13 and E14 days post-coitus, respectively). Grafts derived from 4 mm embryos survived poorly, with less than 1% of the implanted dopamine cells surviving. Grafts derived from 9 mm and 10.5 mm embryos were similar to those seen in previous experiments with survival rates of 8% and 7% respectively. The best survival was seen in the group that received 6 mm grafts, which were significantly larger than all other graft groups. Mean dopamine cell survival in the 6 mm group (E12) was 36%, an extremely high survival rate for primary, untreated ventral mesencephalic grafts applied as a single placement, and more than fivefold larger than the survival rate observed in the 10.5 mm (E14) group. As E12 ventral mesencephalic tissues contain few, if any, differentiated dopamine cells we conclude that the large numbers of dopamine cells seen in the 6 mm grafts must have differentiated post-implantation. We consider the in vivo conditions which allow this differentiation to occur, and the implications for the future of clinical trials based on dopamine cell replacement therapy.