Very little data are available from population-based studies on congenital hypothyroidism (CH) epidemiology and patterns of associated birth defects. By linking data from two population-based registries, we describe the epidemiology of CH and associated defects in Atlanta from 1979-1992. Cases included all infants with CH born from 1979-1992 to mothers residing in the metropolitan Atlanta area at the time of birth. We ascertained CH cases by reviewing newborn screening records and records of the Metropolitan Atlanta Congenital Defects Program (MACDP), a population-based registry of all serious birth defects diagnosed during a child's first year of life. We linked CH cases with MACDP records to ascertain the presence of serious birth defects among infants with CH. Of 97 infants identified with CH through newborn screening and/or MACDP (1:5,000 live births), 87 had primary CH and 10 had secondary. The rate of primary CH was higher among nonhispanic whites than among blacks (1:4,400 vs. 1:10,000) and among females compared with males (1:4,000 vs. 1:7,700). Among infants with primary CH, 77 had isolated CH, 3 had Down syndrome, and 7 had unrelated major structural defects. Based on Atlanta population rates of Down syndrome and major structural anomalies, we infer i) infants with Down syndrome have a 35-fold increased risk for primary CH compared with infants in the general population (P < .0001); ii) infants with primary CH have a 2.2-fold increased risk for major structural anomalies (P < .05). Because this is the first population study of CH in the United States in which data from two population-based registries were linked, the epidemiologic pat-terns and associated defects are more representative than those found in studies based on newborn screening records only.