Background N-terminal pro-B type natriuretic peptide (NTpBNP) is a potential marker of cardiac failure. Methods The Roche ElecsysR(TM) 1010 and 2010 assays for NTpBNP were 2 evaluated for precision, sample stability, and correlation between sample types and with other natriuretic peptides. Samples from 290 individuals aged 45-89 years with no cardiovascular risk factors, renal failure, electrocardiogram changes, evidence of structural abnormalities, or wall motion abnormalities on echocardiography and with an ejection fraction >50% were used to provide reference NTpBNP ranges. Results The intra-assay imprecision was < 10% across the analytical range and < 3% at all concentrations analysed > 30 ng/L. Inter-assay imprecision was 5.36.7% on the Elecsys 1010 and 4.4-5.0% on the Elecsys 2010, in the range 380-13000ng/L. There was no statistically significant change in NTpBNP following storage in whole-blood samples at room temperature for 24 h before centrifugation; serum samples at room temperature for 7 days, at 4 degreesC for up to 11 days on clotactivation gel or 22 days separated from the gel. NTpBNP concentrations were stable throughout five freeze-thaw cycles. There was a close correlation between NTpBNP concentrations in matched serum, EDTA plasma and lithium-heparin plasma samples. NTpBNP and BNP were more closely associated than were N-terminal proatrial natriuretic peptide and NTpBNP. This association was stronger at lower concentrations. NTpBNP concentrations increased with age, with values higher in women than men. Conclusions NTpBNP is a stable molecule that can be measured easily and precisely using the Roche Elecsys 1010 or 2010 immunoassay analysers.