Serum amyloid A1 (SAA1), one of the two isotypes of acute phase SAA, is the predominant precursor to amyloid A (AA) protein, the chief constituent of fibrillar deposits in reactive (AA) amyloidosis. Prolonged hyperexpression of SAA protein accompanying chronic inflammation is critical to, but seems not to be sufficient for, the development of AA amyloidosis. Several previous studies have investigated the possibility of linkage between SAA1 exon 3 polymorphisms and susceptibility to amyloidosis. While the SAA1.1 allele was found to have a negative association with amylodosis in Japanese subjects, it showed a positive association in Caucasians. Moriguchi and colleagues recently showed that a single nucleotide polymorphism (SNP) at position -13 in the SAA1 5' flanking region was more strongly associated with amyloidosis than was the exon 3 polymorphism. To test whether this SNP may be an amyloidogenic factor common to Japanese and Caucasians, we have analyzed the SAA1 gene in amyloid and non-amyloid patients of both ethnic groups for the presence of T or C at position -13 and for exon 3 polymorphisms (SAA1.1, 1.3 or 1.5). The frequency of the -13T allele was 0.708 and 0.521 in Japanese rheumatoid arthritis patients with and patients without AA amyloidosis, respectively, and 0.536 and 0.196 in American Caucasian patients with AA amyloidosis and control subjects, respectively. In Caucasians, the -13T allele had a stronger association with amyloidosis than did the SAA1.1 allele. These findings suggest that -13T is a genetic background for AA amyloidosis in both Japanese and Caucasians and the difference in prevalence of AA amyloidosis in the two ethnic groups may be due, at least in part, to a difference in the frequency of the -13T SAA1 allele.