The aim of this study is to present an evaluation of HbA(1c) Assay on Architect CI8200 (Abbott Diagnostic). The measurement includes Hb assay by colorimetry and HbA(1c) by immunoturbidimetry. The percentage of HbA(1c) is the report HbA(1c)/complete Hb with a conversion coefficient. Repetability (n = 30; CV: 1.15-1.91%) and reproductibility (n = 30; CV: 2.09-2.64%) are good. Abbott results cannot be returned above 12%. Comparison between HbA(1c) Abbott and HbA(1c) Bio-Rad is performed on 161 patients samples ranging from 4.7 and 12%. Results show a correlation coefficient of 0.9847 (N = 161) with a regression equation: [Abbott] = 1.02x [Bio-Rad]-0,636]. Differences between two methods are normally distributed. 95% of differences lie between limits (-0.61%; +0.61%). Such differences are clinically important and interchangability of two measurements can't be possible now because lack of agreement. We hope that IFCC standardization will reduce these differences. Presence of a jaundice and carbamylation of haemoglobin do not interfer with Abbott assay. Hemoglobin variants are not detected. Therefore, monitoring of diabetic patients with HbA(1c) is possible only if hemoglobinopathy has been identified before.