OBJECTIVES: To determine the type and frequency of thyroid disorders detected in infants with low thyroxine (T4) and nonelevated thyroid-stimulating hormone (TSH) screening test results in the Northwest Regional Newborn Screening Program (NWRNSP) over the 20-year period from May 1975 to May 1995 and to determine the effect of follow-up of these infants on the overall recall rate. STUDY DESIGN: The NWRNSP requests a serum specimen in infants with an absolute T4 level < 38.6 nmol/L (< 3 mg/dl) and in infants with two filter paper T4 concentrations less than the 3%, regardless of the TSH concentration. We conducted a retrospective analysis of infants who were followed up because of low T4 and nonelevated TSH concentrations on newborn screening. To determine the effect of follow-up of infants with low T4 levels, nonelevated TSH concentrations on the recall rate, we selected 1 year (1994) for review. Serum sample requests were evaluated to determine the reason for the request. RESULTS: Over this 20-year period, the NWRNSP detected 450 infants with primary hypothyroidism among 1,747,805 infants screened (1:3,884). Of these, 416 were detected on the basis of low T4 levels and nonelevated TSH screening test results, whereas an additional 34 infants with primary hypothyroidism and 29 infants with hypopituitary hypothyroidism were detected as a result of follow-up of low T4 levels and nonelevated TSH screening test results. This included 25 infants with delayed TSH rise (1:67,226), 9 infants with mild hypothyroidism (TSH levels < 25 mU/L) (1:194,212), 29 infants with hypopituitary hypothyroidism (1:60,269), and 434 infants with T4-binding globulin deficiency (1:4,027). Excluding those with T4-binding globulin deficiency, the false-positive rate was 43.5:1. This compares with an overall false-positive rate of 12:1 for our screening program. CONCLUSION: Follow-up of infants with low T4 and nonelevated TSH concentration on screening led to the detection of 63 additional infants with hypothyroidism, for an overall frequency of 1:27,743. We believe this yield justifies continued follow-up of infants with low T4 levels, nonelevated (TSH) screening test results in our program.