The low specificity of cord serum T4 levels for detecting hypothyroidism, the need to reduce recall rates and the availability of a reliable and quick assay system led to this study designed to determine the advantages of using cord serum free thyroxine levels over total thyroxine levels as a supplement to cord thyroid stimulating hormone (TSH) determination. Sixty-eight out of 75 newborns with a cord TSH > 23 mU/l had both free thyroxine (fT4) and T4 levels measured in the cord serum. All were recalled within the first month of life for reevaluation of their thyroid status. In the majority of cases (46), the fT4 and T4 values corresponded. Hypothyroidism was diagnosed in eight cases; both fT4 and T4 were below the mean in five cases and above the mean in one case; only T4 was below the mean in one and in the remaining one, only fT4 was below the mean. Recall rates were 0.9% with TSH alone and 0.7% when either a fT4 or T4 level that was less than 1 sd above the mean was used as a supplement to TSH. In this cohort, using the fT4 levels instead of the T4 levels brought no change to the specificity or the sensitivity of the screening. The diagnostic sensitivity would have dropped from 100% to 75% if either T4 or fT4 values below the mean were used as a cut-off point for recall.