BACKGROUND: Good glycaemic control in type 2 diabetes can reduce both morbidity and mortality, and monitoring of glycated haemoglobin A(1c) (HbA(1c)) is currently recommended for this purpose. METHODS: Haemoglobin A(1c) can be measured using the portable Bayer DCA 2000+. As part of a community-based screening project for chronic diseases, subjects with impaired fasting glucose concentrations or diabetes had their HbA(1c) concentration measured on the analyser. HbA(1c) measurements were also made in a laboratory using cation-exchange high-performance liquid chromatography. Results were then compared to those obtained in the field. HbA(1c) values were log(e) transformed to obtain a normal distribution. RESULTS: Mean (95% confidence interval) HbA(1c) measured on the DCA 2000+ (n = 39) was 6.3% (5.8, 6.9%), while that measured in the laboratory was 6.5% (6.0, 7.0%). The correlation coefficient (r) between the measurements was 0.96 (P < 0.01). Bland-Altman analysis revealed that the DCA 2000+ estimate of HbA(1c) could be used interchangeably with that from the laboratory (mean bias = 0.1%, limits of agreement - 1.1, 0.8%). The sensitivity, specificity and positive predictive value of the DCA 2000+ estimate of HbA(1c) using a cut-off value of <7% for adequate glycaemic control, were all 100%. CONCLUSIONS: The DCA 2000+ may be used to measure HbA(1c) in remote communities.