GH has previously been shown to be present in peripheral extrapituitary tissues of chick embryos, but the cellular distribution of GH immunoreactivity is still uncertain because of differing immunohistochemical findings. The possibility that this uncertainty reflects differences in fixation of the embryonic tissues was assessed by comparing GH immunoreactivity in tissues fixed in 4% (w/v) paraformaldehyde or Carnoy's fluid (60% ethanol (v/v); 30% chloroform (v/v); 10% acetic acid (v/v)).A widespread distribution of GH immunoreactivity was seen in paraformaldehyde-fixed tissues, although it was particularly intense in the spinal cord, dorsal and ventral root ganglia, notochord, myotome, epidermis, crop, heart, lung and humerus. In marked contrast, GH immunoreactivity in embryonic tissues fixed with Carnoy's was more discrete and mainly restricted to marginal and mantle layers of the spinal cord, spinal nerves, the ventral root ganglia and the extensor nerve of the anterior limb bud. Since these are neural derivatives, Carnoy's fixation appears to preferentially result in neural GH staining, whereas GH staining in neural and non-neural tissues is seen after paraformaldehyde fixation. Carnoy's, because it is a precipitive fixative, may only fix large GH moieties, whereas GH in peripheral tissues includes numerous molecular variants, many of which are of relatively small size. Paraformaldehyde, because it is a cross-linking fixative, preferentially fixes peptides and small proteins, and it may therefore fix more GH moieties than Carnoy's fluid. Carnoy's fixation appears to underestimate GH immunoreactivity in immunohistochemical studies on the cellular distribution of GH-like proteins in embryonic chicks.