Context  The guidelines of the National Academy of Clinical Biochemistry advocated the use of low bloodspot TSH (b‐TSH) threshold for newborn screening of congenital hypothyroidism (CH). The impact generated by the application of this indication is largely unknown.

Objective  To determine the impact on CH epidemiology and classification generated by the introduction of low b‐TSH cutoff.

Design  Retrospective study of 629,042 newborns screened with b‐TSH cutoffs of 12 (years 1999–2002) or 10 mU/l (2003–2005).

Measurements  Congenital hypothyroidism incidence and classification. Results were compared with those virtually obtained with the previous cutoff (20 mU/l). Clinical re‐evaluation after L‐T4 withdrawal of a representative group of 140 CH children at 3–5 years.

Results  Low b‐TSH cutoffs allowed the detection of 435 newborns with confirmed CH (incidence 1:1446). Forty‐five percent of CH infants, including 12/141 dysgenesis, would have been missed using the 20 mU/l cutoff. In contrast to current classification, 32% CH newborns had thyroid dysgenesis and 68% had a gland in situ (GIS). Premature birth was present in 20% of cases being associated with a 3–5 fold increased risk of GIS CH. Re‐evaluation at 3–5 years showed a permanent thyroid dysfunction in 78% of 59 CH toddlers with GIS.

Conclusions  The use of low b‐TSH cutoff allowed the detection of an unsuspected number of children with neonatal hypothyroidism, evolving in mild permanent thyroid dysfunction later in life. The incidence of CH in this Italian population appears to be double than previously thought with a clear‐cut prevalence of functional defects over dysgenetic ones.