Our regional cord blood screening programme for detecting neonatal hypothyroidism using initial cord blood thyroxine (T4) determinations, with supplemental thyrotropin (TSH), and triiodothyronine resin uptake (T3U) measurements, gave an incidence of thyroid abnormalities of 1/3000 births, with 1/5000 infants having severe primary hypothyroidism. No hypothyroid infant detected in the programme had been suspected clinically before the screening and, in retrospect, only a few babies had any signs of hypothyroidism. Supplemental TSH and T3U determinations were required on 8-12% of the population screened initially with a T4 test to avoid missing affected cases. With an initial T4 and supplementary TSH and T3U testing on cord blood serum, recalls to exclude primary hypothyroidism were reduced to 0.16% of the screened population. The incidence of abnormalities detected in this cord blood screening programme was comparable with that reported by others using neonatal dried blood screening methods, indicating that cord blood screening can be effective provided the appropriate recall criteria and transport conditions are used. Nevertheless, for several practical reasons, neonatal dried blood methods are recommended as the screening test of choice for surveying large populations over extensive geographical areas.