High heart rate variability (HRV) has been associated with more efficient autonomic control, allowing more responsivity and sensitivity to changing environmental demands. A number of specific periodicities have been identified in the spectra of cardiac time series. A high frequency component related to respiratory sinus arrhythmia, a low frequency component related to blood pressure variability, and a very low frequency component thought to reflect thermoregulation have been reported in the literature. However, the source of the very low frequency component has not been extensively investigated in humans using non-invasive methods and analytic techniques that do not rely upon stationarity. We investigated HRV in response to both hot and cold thermal challenge in healthy adults using time-frequency analysis. This analytic technique does not rely upon signal stationarity. The results suggest that very low frequency power may reflect thermoregulation to ambient temperature changes. Implications for prediction of cardiac events are discussed.