The heart rate (HR)-minute ventilation (VE) relationship has been shown to be nonlinear and can be expressed as two distinct straight lines. This study is to assess the correlation of the initial HR-VE slope to clinical parameters. Maximum treadmill exercise tests were performed in 100 healthy volunteers (age 19-77 years) using a ramp protocol in which work-rate increases linearly with exercise. Breath-by-breath VO2, VCO2, and VE were measured, and HR and BP were monitored throughout the exercise. The HR-VE curve demonstrated nonlinearity with a breakpoint determined by a change point analysis. This breakpoint was significantly higher than that of the anaerobic threshold. The VE at the HR-VE breakpoint was 56.4 +/- 19.4 and VE at the VE-VO2-VO2 breakpoints were 48.0 +/- 18.3 (P < 0.0001) and 40.1 +/- 16.5 (P < 0.0001), respectively. The HR at this HR-Ve breakpoint was 77.7 +/- 12.9% of the HR range. The first slope, S1 (1.76 +/- 0.64) was steeper than the second slope, S2 (0.66 +/- 0.39). Although there was a gender difference for S1, the best clinical predictor on a stepwise multiple regression analysis was body surface area (BSA) which explained 47% of the variance. It was concluded that nonlinearity of the HR-VE curve can be expressed as two straight lines. The breakpoint is beyond the anaerobic threshold and can be estimated to be approximately 75% of the maximal predicted HR. BSA is the only clinical parameter that significantly predicts the initial slope of the HR-VE curve. This can be of great importance in the programming of rate-adaptive pacemakers using a VE.