The extent to which age-associated changes in aerobic capacity and body composition modulate the age-associated decline in heart rate variability (HRV) is unknown. We therefore measured HRV, peak O2 consumption (VO2peak) during treadmill testing, and relative weight (body mass index; BMI) in 164 healthy normotensive adults (75 men, age 20-87 yr) from the Baltimore Longitudinal Study on Aging. Two components of HRV, respiratory sinus arrhythmia (RSA; 0.12-0.40 Hz) and 0.10-Hz variability (0.06-0.10 Hz), were extracted from 8-min electrocardiogram recordings in the supine, seated, and standing postures. RSA, 0.10-Hz variability, and VO2peak varied inversely with age; BMI was unrelated to age. Age contributed 15.5-21.1% independent variance to RSA and 13.2-17.3% independent variance to 0.10-Hz HRV. VO2peak did not contribute significantly to RSA or 0.10-Hz HRV beyond the effect of age in any posture. There were no consistent independent contributions of BMI to HRV. Thus, in this population-based sample, age-associated changes in aerobic capacity and relative body weight do not provide the primary explanation for the decline in HRV observed with advancing age.