Blood measurements of BNP and NT-proBNP, its catabolite, improve diagnosis for patients admitted to emergency departments with dyspnoea. In this paper, we have compared the BNP to the NT-proBNP for 119 dyspnoeic patients using at random clear clinical status. Among the test group of 119 patients, 57 showed coherent biological results for the 2 markers. These results confirm the final clinical diagnosis. Nine patients with congestive heart failure had abnormally low BNP and NT-proBNP rates. Six of these patients experienced long delays (longer than 48 hours and less than 72 hours) between their admission in emergency and the biological measurement of the natriuretic biomarkers. Three of the other patients could be not only flash OAP cases with a fast growth and a fast normalisation of BNP but also could have existing genetical factors. These genetical factors leading to high variability in BNP synthesis are not related to physiological or cardiac factors. 43 patients showed a mismatch between BNP and NT-proBNP. BNP appeared to be unstable in vitro. The lack of stability in whole blood or plasma samples is increased by sampling in a glass EDTA collection tube and too long delays in transferring the samples from the emergency area and the laboratory in a big hospital. Ten patients showed a mismatch with abnormally high NT-proBNP or false positive results. Among these 10 patients, 5 had renal dysfunction with a high level of creatinine concentration. It is clear that all Diagnostics Manufacturers should now propose different cut-off for natriuretic peptides tests according to the degree of patients' renal impairment.