Experience of neonatal thyroid screening has proved its benefit and efficacy for more than 10 years. Some 35 million infants have been screened so far, and about 9 million infants are screened yearly throughout the industrialised world . Additional preliminary screening programmes have been established in regional areas of developing countries. Two international conferences on Neonatal Thyroid Screening for permanent primary congenital hypothyroidism have indicated a frequency of 1 in 3800-4000 infants worldwide, and of i in 3500 infants in Europe . Permanent congenital hypothyroidism is caused in about 60% by dysplastic or ectopic thyroid glands, and in about 30% by thyroid agenesis. The remaining 10% include thyroid dyshormonogenesis, hypothalamic-pituitary dysfunction and hypothyroidism induced by placental transfer of maternal thyroid antibodies. The last two causes have an incidence of approximately 1 : 100000 each. Thyroid screening is usually performed by primary measurement of TSH in cord blood or in blood drops placed and dried on filter paper cards on the 5th day of life, or of primary T4 in blood drops taken on the 5th day of life supplemented by TSH measurement in infants with low T4 values. With either approach, aimed at detecting primary hypothyroidism with elevated TSH values, congenital hypothyroidism will be missed in some infants. Using TSH-screening, hypothalamiepituitary hypothyroidism and delayed TSH elevation in cases of slowly developing primary hypothyroidism may be missed. In screening programmes using cord blood delayed TSH elevation especially is not detectable, as indicated in the report of Virtanen et al. in this issue. Infants with congenital hypothyroidism induced by placental transfer of maternal thyroid antibodies and homozygotic twins with feto-fetal placental blood exchange may also show TSH and T4 values within normal limits in cord blood but many show an obvious pathological trend on the 5th day of life . Since the predictive value of TSH-screening by filter paper blood spots taken on the 5th day of life is considered to be higher than that of cord blood screening , newborns with late developing hypothyroidism should be detected with more certainty by the widely used screening on the 5th day of life. The advantage of an earlier diagnosis by cord blood screening is generally considered to be smaller than the disadvantage of a possibly higher percentage of missed patients. For this reason and for the great practical advantage of combining thyroid screening with screening for other inborn errors of metabolism (e.g. PKU), these screening programmes were recommended by Burrow and Dussault  and others to be initiated together on the 5th day of life. The higher incidence of congenital hypothyroidism in Finland may be due either to genetic causes or possibly may be artificial. It is well known from neonatal thyroid screening that a great percentage of border-line values normalises not before day 8-10. The observation of low TSH and low T4 concentrations in critically ill infants is usually a physiologically induced regulative process that normalises with convalescence . This applies especially to newborns exposed to iodine during treatment. Preterm infants with neonatal diseases frequently present a status of hypothyroxinaemia that must not be interpreted as transient hypothyroidism. There is no doubt that treatment of permanent congenital hypothyroidism should be initiated as soon as possible. Since the prognosis of mental development according to recent reports is excellent if treatment starts before the age of 1 month, there should be enough time to reach a clear-cut diagnosis in border-line cases also by using the world-wide instituted hypothyroidism screening on the 5th day of life.