INTRODUCTION: Measurement of hemoglobin A1c (HbA1c) is used as an objective measure of long-term blood glucose control in diabetic patients. Recent improvements in automation combined with new recommendations for precision and accuracy have caused us to reevaluate our methods for measuring HbA1c. OBJECTIVE: We evaluated a newly automated high-performance liquid chromatography (HPLC) instrument for measurement of HbA1c (Tosoh A1c 2.2 Plus Glycohemoglobin Analyzer, Tosoh Medics, Foster City, Calif) and compared the results obtained by HPLC to those obtained with an immunoassay (Hitachi 911, Boehringer Mannheim Corporation, Indianapolis, Ind). RESULTS: The Tosoh analyzer was found to be linear in a range of 5.3% to 17% and had a throughput of 20 samples per hour. HbA1c results for 102 patient samples by the 2 techniques showed good correlation, with a slope of 0.87 and an intercept at 1.27% +/- 0.15%. Both the total and within-run coefficients of variation were consistently lower for the HPLC method compared with the immunoassay method. The HPLC method produces a chromatogram that shows the different hemoglobin fractions, allowing identification of abnormal hemoglobin variants. In heterozygous individuals, HbA1c measurements are made with no interference from the hemoglobin variant. In the case of homozygous or doubly heterozygous hemoglobin variants, the Tosoh HPLC identifies the hemoglobin variants as such and correctly does not report a HbA1c value in the presence of a markedly decreased amount of hemoglobin A. CONCLUSIONS: The Tosoh HPLC provides adequate throughput and improved precision, and the method is traceable to the Diabetes Control and Complications Trial.