The embryonic chicken cartilage somatomedin bioassay was modified so that human serum stimulation of simultaneous [3H]methylthymidine and H2[35S]O4 incorporation could be assessed. The assay consisted of a 6 h pre-incubation of 10 day pelvic rudiments in enriched buffer, followed by a 24 h incubation with buffer and low (0.5, 2 and 5% v/v) serum concentrations. Both labels were present for the final 6 h. Other modifications were shortening of washing, elimination of drying and weighing, and simplification of digestion. Normal human serum produced a linear log dose-response with these serum concentrations. Potency ratios in patients with GH deficiency were less than those of normal adults for both thymidine 0.39 +/- 0.05 (mean +/- SEM, n = 16, range 0.22-0.71) vs. 0.90 +/- 0.05 (n = 19, 0.62-1.36, P < 0.001) and for sulphate 0.40 +/- 0.04 (0.15-0.65) vs. 94 +/- 0.05 (0.61-1.29, P < 0.001). Potency ratios for both labels rose following administration of a single dose (0.2 IU/kg im) of hGH to 4 GH deficient children. The reliability of prediction of GH deficiency, reproducibility, and precision were similar to other Sm bioassays. The major advantages of these modifications were the ability to examine 2 cartilage metabolic processes simultaneously and the small amount of serum (350 mul) necessary for patient assays.