In this study we describe the ontogeny of [ 125 I]rat-α -calcitonin gene-related peptide binding sites in the spinal cord of fetal and postnatal sheep. The density and distribution of binding sites has been compared with the localization of calcitonin gene-related peptide like-immunoreactivity at corresponding stages of development [Nitsos I. and Rees S. (1993) Neuroscience 54, 239-252]. At 68 days of gestation (term = 146 days), the earliest fetal tissue examined, there was no evidence of binding sites in lamina I or the outer regions of lamina II (lamina II o ), although there was a sparse distribution of binding sites in the inner region of lamina II (lamina II i ). By comparison, binding appeared to be more marked in laminae III-V and more concentrated again in laminae VI-X. This distribution essentially remained constant until 124 days, when there appeared to be a marked increase in the density of binding sites throughout the gray matter, particularly in the dorsal horn in the lateral extent of both lamina II, and 11; as well as in laminae III, V and VI. This increase was also observed in the intermediate zone (lamina VII) and in lamina X. Binding in the ventral horn, which was diffuse until this stage, now became particularly dense in the medial and lateral regions of the horn. From 124 days to one month postnatal, there was no marked change in the density or distribution of binding sites. In the adult, binding density decreased slightly throughout the spinal cord with the exception of laminae V-VII and lamina X, where the densities were similar to those seen in the neonate. Binding sites were not always coincident with the distribution ofimmunoreactivity for calcitonin gene-related peptide. This was particularly evident in lamina I and the medial aspect of lamina II o , where there was an intense expression of calcitonin gene-related peptide but an absence of binding sites throughout gestation, in laminae III and IV, where binding sites were present and immunoreactivity absent, and in the ventral hom up until 104 days, where there was a concentration of binding sites in the neuropile but no evidence of immunoreactivity for calcitonin gene-related peptide. In the white matter, prominent binding was associated with the ventral funiculus at 68 days. Binding extended into the lateral funiculus from 77 to 104 days, after which it was largely restricted to the dorsolateral fasciculus. Thus, we have shown that binding sites for calcitonin gene-related peptide are present well before mid-gestation in the sheep spinal cord and that differences exist between the localization of binding sites and immunoreactivity for calcitonin gene-related peptide.