Aims: To evaluate and compare the recall rate in congenital hypothyroidism screening project in Isfahan, first using an approach involving measures of both TSH and T4 and then using TSH alone. Methods: From June 2002 to January 2005, serum TSH and T4 level of referred neonates were measured at 3rd to 7th day of birth through venous sampling. If neonates’ serum TSH was >20 mIU/l or T4 was <6.5 µg/dl by the first protocol, or TSH was >20 mIU/l by the second protocol, they were recalled. TSH and T4 were measured using an immunoradiometric assay and radioimmunoassay, respectively. Neonates with TSH > 10 and T4 < 6.5 on their second measurement were considered as congenitally hypothyroid. Results: Serum T4 and TSH of 29,425 neonates by first and 57,235 neonates by second recall approach were measured. Recall rate was higher in the first protocol (2.2% vs. 0.6%, p < 0.05). Most of the recalled neonates in the first protocol were recalled for low T4 level (p < 0.05). The prevalence of CH was 1 in 350 livebirths. Conclusion: Although the recall rate was in the acceptable range by either approach, the TSH alone protocol seems to be a more sensitive and practical approach with the least recall burden and considering the high prevalence of CH in our region merit adaptation of widespread screening for CH using TSH measurements from heel stab blood spotted on filter paper.