BACKGROUND: Misalignment of circadian systems is detrimental to human health. However, only a few studies have examined the influence of late meals on the human circadian system. The purpose of this study was to investigate whether delayed meals affect circadian rhythm as assessed by heart rate variability (HRV). METHODS: Seven healthy men (aged 22.4+/-0.4years) participated in this study, which involved delaying mealtimes by 5h for two weeks. Prior to this study, the regular mealtimes of participants were at 08:00, 13:00, and 18:00 (Baseline). During the intervention, each meal was provided at 13:00, 18:00 and 23:00 (Late mealtimes). Circadian variation was assessed by HRV based on R-R intervals calculated from a pulse at the rising phase of each spike in the QRS complex in a 24-h electrocardiograph record. The ratio of low frequency (LF) to high frequency (HF) power (LF/HF) and the ratio of HF power to total power (%HF) were calculated using spectral analysis. The phase and amplitude of HRV variables in a 24-h period were mathematically obtained with double cosinor analysis. RESULTS: The acrophase and amplitude for LF power, HF power and %HF in a 24-h period were not significantly different between Baseline and Late mealtimes. On the other hand, the acrophase significantly differed between Baseline and Late mealtimes for heart rate (95%CI, 1.1-3.8h), standard deviation of R-R intervals (95%CI, 2.5-4.4h) and LF/HF (95%CI, 1.1-2.3h). CONCLUSIONS: Our findings suggest that delayed mealtimes shift the phase of circadian rhythm of the autonomic nervous system.