The thyrotropin receptor (TSH-R) gene is a candidate for genetic susceptibility to Graves' disease (GD). Previous case control studies investigating allelic association of a polymorphism at position 253 (C253 to A253) of the TSH-R gene have shown conflicting results. We genotyped two independent case control datasets (UK Caucasian and Hong Kong Chinese), for the A253 polymorphism. The Transmission Disequilibrium Test was also used in a third family-based dataset that included 89 UK Caucasian families (both parents, a GD sibling and an unaffected sibling). Genotyping was performed by polymerase chain reaction (PCR)-amplification of genomic DNA and Tth111 I restriction enzyme digestion. No difference in frequencies of the A253 polymorphism between GD (21/204, 10.3%) and controls (34/358, 9.5%) was found in the UK Caucasians (chi2 = 0.093; p = NS). A similar finding was observed in GD (0/96, 0%) and controls (2/71, 2.8%) in Hong Kong Chinese subjects (chi2 = 2.73; p = NS). Results from the 89 UK families showed no deviation from the expected transmission frequency of 0.5, from parents heterozygous for the A253 allele to either Graves' or unaffected offspring (Fisher's exact test p = 0.22) and, therefore, confirmed a lack of evidence of linkage disequilibrium between the A253 allele and GD.