Many assays 1(st), 2(nd) even 3(rd) generation are at present available to determine the concentration of cardiac troponin I and T. With the redefinition of upper reference value in the acute coronary syndromes, the aim of this study was to evaluate the clinical and analytical performance of 2 troponins assays: Troponin Ic 2(nd) generation (AccuTnI) on Access 2 of Beckman Coulter and Troponin Tc 3(rd)generation (Troponin T STAT) on Elecsys 2010 of Roche Diagnostics. The analytical performance observed with these 2 assays are accurate (analytical and functional sensitivity, repetability and reproductibility). Comparing each method with Dade Behring assay (Flex Troponine-I Cardiaque, TROP) on Dimension RxL, the correlation observed with AccuTnI kit on Access 2 can be put into the equation: AccuTnI = 1.08 (TnIc TROP) - 0.34, r = 0.99. On the contrary, it's more difficult to compare cTnI and cTnT. The study of decisonnal values indicated by Beckman Coulter for cTnI (0.04 microg/L at the 99 degrees percentil, 0.06 microg/L for a CV < or =10%) show a better specificity (76%) and predictive positive value (89%) with a sensitivity at 100% at 0.1 microg/L, fixed and used in the laboratory for its better agreement between sensibility / specificity and its imprecision below 10 %. For the cTnT values published by Roche Diagnostics (0.01 microg/L), at the 99 degrees percentil and 0.03 microg/L for a CV < or = 10%, the specificity is lower, so the decisionnal value 0.1 microg/L seems to be more suitable. During this study, few false positive and negative cTnT values have been observed, in patients with complex pathologies; this eventuality must be taken in consideration if clinical findings are not in good accordance with laboratory results.