The aim of this work was to test the differentiation capacities of neural crest-derived cells that had already migrated to the gut and were in the process of forming the intrinsic gut plexuses, when they were withdrawn from the gut wall environment. The experiments were performed on the quail bowel of embryos at 4 to 12 days of incubation (E4 to E12), which was dissociated into a single cell suspension. The crest cells invading the gut carry the NC1/HNK1 epitope. This allowed us to select them as singlets under the control of a uv-illuminated inverted microscope. Crest-derived cells from the gizzard were then clonally cultured on a feeder layer of growth-inhibited 3T3 fibroblasts, as described before for cephalic and truncal neural crest cells. The clonal efficiency of these cells reached about 28% from E4 to E7 and decreased sharply at E8. The size of the clones generated by cells taken between E4 to E8 varied considerably. Large clones (> 10(3)) were numerous from E4 and E5 gizzard-derived cells, whereas E7 and E8 cells produced only very small colonies. The phenotypic diversity of the clones decreased similarly during the time period under scrutiny. Two phenotypes never encountered in the enteric plexuses were found in the cultures: adrenergic cells and glial cells expressing SMP, a marker exclusively expressed by Schwann cells in vivo and normally absent in enteric glia. The capacity to yield adrenergic cells was present up to E6 in clonal cultures of gizzard neural crest-derived cells. Mass cultures of cells dissociated from the gizzard and bowel, including gut mesenchymal and epithelial cells in addition to cells of neural crest origin, revealed that both adrenergic and SMP+ cells could arise under these conditions even in a defined culture medium devoid of serum and chick embryo extract. The number of TH+ cells developing in bowel cultures reached a peak at E7, while at this stage no TH+ cells could be obtained from gizzard cultures. A craniocaudal gradient of disappearance of adrenergic precursors in the gut could thus be demonstrated. Finally we point out that these and previous experiments from other laboratories show that precursor cells of the sympathoadrenal lineage are present in all types of avian sensory and autonomic ganglia during development even though, in most of them, these cells never express the adrenergic phenotype.