1. The dorsal horn (DH) of chickens exhibits a novel pattern of cytoarchitectonic lamination among vertebrates, whereby lamina III lies medial, rather than ventral, to lamina II. Indeed, cutaneous nerves labeled with horseradish peroxidase (HRP) form two separate projections across the mediolateral axis of the superficial DH; each projection is somatotopically organized, such that two non-overlapping somatotopic maps are formed: the medial map within lamina III and the lateral map within lamina II. Interestingly, these two projections of cutaneous nerves are differentially labeled by HRP ligands. The present experiments were designed to address whether the separate subpopulations of cutaneous afferents, as demonstrated neuroanatomically, also differ physiologically on the basis of myelination, fiber diameter, and/or sensory modality. 2. Extracellular multi- and single-unit recordings were obtained in the DH at the ninth synsacral level of spinal adult chickens anesthetized with alpha-chloralose. Activity in lateral (laminae I/II) and medial DH (laminae III/IV) was studied after both electrical stimulation of the caudal femoral cutaneous nerve (CFC) and natural stimulation of the skin. Single units were characterized in terms of the conduction velocity (CV) and sensory modality of their cutaneous afferent inputs. 3. In multiunit recordings, electrical stimulation of the CFC above C-fiber intensities elicited a robust, long-latency response in lateral (laminae I/II) but not medial DH (laminae III/V). The afferents responsible for this late lateral response were C-fibers, as evidenced by a CV of approximately 1 m/s; birds were spinalized to rule out long spinal loops. In contrast, only a single, short-latency (and low-threshold) multiunit response was seen in medial DH, even after activation of C-fibers. 4. Natural stimulation of skin revealed a clear segregation of sensory modalities between medial (laminae III/IV) and lateral DH (laminae I/II). Innocuous mechanical stimuli were extremely effective at eliciting multiunit activity in medial DH, but ineffective in lateral DH. In contrast, noxious mechanical and thermal stimuli were extremely effective at eliciting multiunit activity in lateral DH, but were ineffective in medial DH. 5. In single-unit studies, primary afferent inputs to units in medial DH (laminae III/IV) had an average CV close to 43 m/s; no medial units received exclusive inputs from afferents with CVs less than 5.5 m/s. In contrast, primary afferent inputs to units in lateral DH (laminae I/II) had an average CV close to 10 m/s; 20% of the lateral units received exclusive inputs from C-fibers (CVs less than 1.7 m/s; N = 1.2 m/s).(ABSTRACT TRUNCATED AT 250 WORDS)