Algorithms for treatment of diffuse bleeding in cardiac surgery are based on intervention thresholds of coagulation tests, such as rotational thromboelastometry (ROTEM) or conventional laboratory tests. The relationship between these two approaches is unclear in patients with increased risk of coagulation abnormalities. We retrospectively analyzed the data of 248 patients undergoing major cardiac and/or aortic surgery. ROTEM and conventional laboratory tests were performed simultaneously after termination of cardiopulmonary bypass and protamine administration to investigate the extrinsic and intrinsic system, and to determine deficiencies in platelets and fibrinogen. We evaluated the association between ROTEM and conventional tests by linear regression analysis and compared the frequency of exceeding established thresholds for clinical intervention. Significant linear associations between ROTEM 10 min after the start of coagulation, and plasma fibrinogen concentration or platelet count (FIBTEM A10, R² = 0.67, p < 0.001; EXTEM A10, R² = 0.47, p < 0.001) were obtained. However, the 95% prediction intervals exceeded clinically useful ranges (92–233 mg/dL fibrinogen at the intervention threshold of FIBTEM A10 = 10 mm; 14 × 10³–122 × 10³/µL platelets at the intervention threshold of EXTEM A10 = 40 mm). The association between EXTEM and INR (R² = 0.23), and INTEM and aPTT (R² = 0.095) was poor. The frequency of exceeding intervention thresholds and, consequently, of triggering treatment, varied markedly between ROTEM and conventional tests (p < 0.001 for all comparisons). The predictability of conventional coagulation test results by ROTEM is limited, thus hampering the interchangeability of methods in clinical practice.