Plasma concentrations of N-terminal pro-brain natriuretic peptide (NT-proBNP) are raised in patients with left ventricular dysfunction. Measurement of this peptide has a potential diagnostic role in the identification and assessment of patients with heart failure. The stability of this peptide over time periods and conditions pertaining to routine clinical practice has not been reported previously. Blood samples were obtained from 15 subjects. One aliquot was processed immediately, and the remaining portions of the blood samples were stored for 24 h or 48 h at room temperature or on ice prior to processing. Plasma concentrations of NT-proBNP were measured with a novel immunoluminometric assay developed within our laboratory. Mean plasma concentrations of NT-proBNP were not significantly different whether blood samples were centrifuged immediately and stored at -70 degrees C or kept at room temperature or on ice for 24 h or 48 h. The mean percentage differences from baseline (reference standard) were +5.2% (95% confidence interval +18.2 to -7.8%) and +0.8% (+15.2 to -13.7%) after storage for 24 h at room temperature or on ice respectively, and +8.9% (+24.2 to -6. 5%) and +3.2% (+15.1 to -0.9%) for storage for 48 h at room temperature or on ice respectively. Pearson correlation coefficients for baseline NT-proBNP concentrations compared with levels at 48 h at room temperature or on ice were r=0.89 and r=0.83 respectively (both P<0.0001). Thus NT-proBNP extracted from plasma samples treated with EDTA and aprotinin is stable under conditions relevant to clinical practice.