BACKGROUND: Mental stress (MS) reduces diastolic function (DF) and may lead to congestive heart failure with preserved systolic function. Whether brain natriuretic hormone (brain natriuretic peptide [BNP]) mediates the relationship of MS with DF is unknown. METHODS: One hundred sixty individuals aged 30 to 50 years underwent 2-hour protocol of 40-minute rest, videogame stressor and recovery. Hemodynamics, pro-BNP samples and DF indices were obtained throughout the protocol. Separate regression analyses were conducted using rest and stress E/A, E and E/E as dependent variables. Predictor variables were entered into the stepwise regression models in a hierarchical fashion. At the first level, age, sex, race, height, body mass index, pro-BNP and left ventricular mass (LVM) were permitted to enter the models. The second level consisted of systolic blood pressure (SBP), diastolic blood pressure (DBP) and heart rate (HR). The final level contained cross-product terms of race by SBP, DBP and HR. RESULTS: E/A ratio was lower during stress compared to rest and recovery (P < 0.01). Resting E/A ratio was predicted by a regression model of age (-0.31), pro-BNP (0.16), HR (-0.40) and DBP (-0.23) with an R² = 0.33. Stress E/A ratio was predicted by age (-0.24), pro-BNP (0.08), HR (-0.38) and SBP (-0.21) with total R² = 0.22. Resting E model consisted of age (-0.22), pro-BNP (0.26), DBP (-0.27) and LVM (-0.15) with an R² = 0.29. Stress E was predicted by age (-0.18), pro-BNP (0.35) and LVM (-0.18) with an R² = 0.18. Resting E/E was predicted by race (0.17, B > W) and DBP (0.24) with an R² = 0.10. Stress E/E consisted of pro-BNP (-0.36), height (-0.26) and HR (-0.21) with an R² = 0.15. CONCLUSIONS: pro-BNP predicts both resting and stress DF, suggesting that lower BNP during MS may be a marker of diastolic dysfunction in apparently healthy individuals.