Résumé : Background: Early clot amplitudes measured on thromboelastometry (ROTEM®) predict maximum clot firmness (MCF) in adults. In this multicentre, retrospective study, we aimed to confirm the suspected relationship between early ROTEM® variables and MCF, in children undergoing cardiac or non-cardiac surgery. Methods: 4762 ROTEM® tests (e.g. EXTEM, INTEM, FIBTEM, APTEM, and HEPTEM) performed in children undergoing cardiac or non-cardiac surgery at three University hospitals between January 2011 and June 2014 were reviewed. To assess the correlation between clot amplitudes measured after 5, 10 and 15 min and MCF, each variable was compared with the corresponding MCF by calculating Spearman's correlation coefficient. Results: For the EXTEM® test, we observed that amplitude measured after 5 min (A5: r=0.91, P<0.001), 10 min (A10: r=0.95, P<0.001) and 15 min (A15: r=0.96, P<0.001) were strongly correlated to MCF. The same correlations were observed for INTEM® test (A5: r=0.93, P<0.001; A10: r=0.97, P<0.001; A15: r=0.97, P<0.001), and FIBTEM® test (A5: r=0.93, P<0.001; A10: r=0.94, P<0.001; A15: r=0.96, P<0.001). In addition, the amplitudes measured after five, 10 and 15 min were also strongly correlated with MCF in the APTEM® and the HEPTEM® tests. Receiver operating characteristics (ROC) analysis confirmed that A5, A10, A15 strongly predicted decreased MCF on all ROTEM® tests. Conclusions: This study confirmed that early values of clot amplitudes measured as soon as five, 10 or 15 min after clotting time could be used to predict maximum clot firmness in all ROTEM® tests.