The recommended method for assessing long-term blood glucose control in diabetic patients is the measurement of glycated haemoglobin (Hb). The Ames DCA 2000 system for assaying glycated Hb uses an immunoassay with a monoclonal antibody specific for an aminoacid sequence within the HblAc molecule. This study compared the performance of the DCA 2000 system for HblAc measurement with that of high-performance liquid chromatography (HPLC). A total of 1.016 insulin-dependent and non-insulin-dependent diabetic patients from 5 outpatient clinics took part. The correlation coefficients between DCA 2000 and HPLC data ranged between 0.94 and 0.98, depending on site. The mean variations and 95% confidence intervals for the differences between the results for each sample were: site A 0.172 (-1.186 to 1.53), site B -0.275 (-1.317 to 0.767), site C -0.146 (-0.868 to 0.576), site D -0.088 (-0.864 to 0.688), and site E -0.251 (-1.099 to 0.597). The sensitivity of the DCA 2000 assay ranged between 80 and 94%, and the specificity between 88 and 100%, depending on site. For pooled results, the correlation coefficient assayed by the two methods was 0.95. The mean variation was -0.116 and the 95% confidence interval -1.23 to 0.998. The sensitivity of DCA 2000 was 91%, and the specificity 94%. DCA tended to underestimate HbAlc slightly as compared to HPLC. This study confirms the reliability of DCA 2000 for measuring glycated Hb. The system is easy to use and provides valuable information for the care of the diabetic patients.