A compelling body of evidence indicates that the suprachiasmatic nucleus (SCN) of the hypothalamus is a pacemaker in the rodent circadian timing system. Two important components of this evidence are studies showing that SCN lesions abolish circadian rhythms and others demonstrating restoration of circadian activity rhythms after transplantation of fetal SCN into the brains of arrhythmic hosts. In the present study, we evaluated what has remained a persisting issue in this transplant literature, the extent to which the exact localization and organization of the transplants is critical to their capacity to restore circadian function in the hamster. The data obtained indicate that the location of the graft in the ventricular system is not crucial to outcome. Grafts in the lateral ventricle, dorsal third ventricle, interventricular foramen, and caudal third ventricle are as capable of restoring circadian function as ones placed in the ventral third ventricle in the vicinity of the lesion. Restoration of rhythmicity does require that the grafts contain a minimum volume of SCN-like tissue as defined by cytoarchitecture and the presence of vasopressin--and vasoactive intestinal polypeptide (VIP)--immunoreactive cells and fibers. There is also an indication that VIP-immunoreactive elements are the component critical to functional recovery. Connections between graft and host are evident in the immunohistochemical material but are quite variable in extent and often very limited. Thus, the data obtained in this study are consistent with the view that restoration of circadian function by fetal grafts requires the presence of SCN, and probably VIP-containing neurons, but does not depend upon the exact location of the graft or the presence of specific connections between graft and host.