The aim of this study was to conduct a longitudinal follow-up on age-related changes in 24-hour total heart beats (THBs) and total premature beats and their correlations in healthy elderly subjects. In 15 healthy elderly subjects (mean age, 70.0 +/- 4.1, age range at 1st recording, 64 to 80 years, 10 females, 5 males), we conducted Holter monitoring twice at an interval of 15 years and analysed age-related changes in THBs, atrial premature beats (APBs), and ventricular premature beats (VPBs), as well as their correlations. The results indicated that THBs, APBs, and VPBs all significantly increased with age in the healthy elderly subjects at a mean age of 70.0 +/- 4.1 (THB: 91074.1 +/- 11515.3 versus 99457.5 +/- 12131.0; P = 0.0004, APB:119.2 +/- 97.8 versus 884.4 +/- 1193.8; P = 0.0008, VPB: 15.2 +/- 53.6 versus 140.7 +/- 228.9; P = 0.0328). Moreover, we divided the subjects into increase and nonincrease groups based on the age-related changes in APB and VPB for 15 years ([n]; Inc-APB: Noninc-APB = 6 : 9, Inc-VPB: Noninc-VPB = 5 : 10). In the increase groups, premature beats tended to increase in proportion to changes in THBs with age (APB: Y = 207.488 + 0.136 X, r = 0.848, P = 0.0303; VPB: Y = -27.594 + 0.028 X, r = 0.727, P = 0.1921). In conclusion, this 15-year follow-up of Holter recordings in healthy elderly subjects revealed that THBs, APBs, and VPBs increased with age, and that the increases in premature beats, especially APBs, were in proportion to those in THBs.