It has recently been shown that mutations in BETA2/NeuroD1 are responsible for the development of type 2 diabetes mellitus (T2DM) in Caucasians. This gene is located near the IDDM7 region and one of its amino acid polymorphisms, Ala45Thr, has been associated with type 1 diabetes (T1DM) in Japanese and Danish populations. The aim of our study is to examine Ala45Thr for its role in T1DM in Caucasians. We used both population-based case-control analysis and family-based transmission/disequilibrium testing (TDT). Genotyping was carried out by the dot-blotting method using P32. Study subjects comprised 202 type 1 diabetes cases (mean age at diagnosis: 11.1 years, mean age at examination: 36.4 years) and 139 controls with normal fasting glucose. For the TDT study, allelic transmission was evaluated in 209 case family trios. The frequency of the Ala45 allele was 70.3 % in cases and 62.9 % in controls (p=0.04), and 47.5 % of cases were Ala45 homozygotes compared to 36.0 % of controls (p=0.03). The TDT component of the study did not achieve statistical significance. However, given the high frequency of this variant even among controls, exceptionally large data sets are needed to provide adequate power for this approach. Our case-control study suggests that the Ala45 variant of BETA2/NeuroD1 may be associated with T1DM in Caucasians (or in linkage disequilibrium with a causative variant). However, this finding should be confirmed by a much larger family-based study.