Background Amino terminal pro-brain natriuretic peptide (NT-proBNP) measurement can detect and assess heart failure. However, compared with traditional clinical parameters, its value in predicting the in-hospital mortality of patients with suspected heart disease has not been reported. Methods We examined the ability of 11 continuous and 21 categorical variables, including NT-proBNP levels measured at the time of admission, to predict in-hospital mortality. The setting was a small Irish rural hospital where 342 consecutive patients with suspected heart disease were admitted as acute medical emergencies. Results The 31 patients who died while in hospital had significantly higher NT-proBNP levels on admission than patients discharged alive (11,548 ± 13,531 vs. 3805 ± 6914 pg/mL, p < 0.0001). Patients who died in-hospital were older, had significantly higher white cell counts, blood urea and modified early warning (MEW) scores, and lower temperatures, blood pressures and oxygen saturation. Four variables were found to be independent predictors of in-hospital mortality: a systolic blood pressure equal to or below 100 mm Hg, a urea level above 13 mmol/L, a white cell count greater than 13 * 10 9/L and a NT-proBNP level greater than or equal to 11,500 pg/mL. The presence of three of these variables was associated with an in-hospital mortality rate of 54%. Conclusions Four variables (i.e. hypotension, elevated urea, leukocytosis and elevated NT-proBNP levels) are comparable independent predictors of in-hospital mortality.