Increased serum leptin concentration has been linked to increased ventilation in patients with mild heart failure (HF). However, in animal models the absence of leptin has also been associated with increased ventilation. This study evaluated the relationship of circulating leptin concentration with exercise ventilation in HF patients.Fifty-eight consecutive ambulatory HF patients were stratified by quintiles of leptin concentration, with a lowest quintile of mean leptin concentration of 1.8 ± 8.9 ng/mL and a highest of 33.3 ± 30.3 ng/mL. Peak exercise ventilatory efficiency (VE/VCO 2 ) was significantly elevated in the lowest (46 ± 6 vs 34 ± 4; P < .01) as well as in the highest (38 ± 8 vs 34 ± 4; P < .05) leptin concentration quintiles compared with the reference middle quintile. Multiple regression analysis adjusted for confounders such as age, sex, and body mass index showed leptin concentration to be independently inversely correlated to VE/VCO 2 in the low-to-normal quintiles (β = −0.64; P < .01), positively in the normal-to-high quintiles (β = 0.52; P = .02), and positively correlated to P ET CO 2 in the low-to-normal quintiles (β = 0.59; P = .01) and inversely in the normal-to-high quintiles (β = −0.53; P = .02).In HF patients, both high and low leptin concentrations are associated with increased VE/VCO 2 and decreased P ET CO 2 with a nonlinear U-shaped relationship, suggesting that either leptin deficiency or leptin resistance may modulate ventilatory control in HF patients.