We read with interest the recent study by Lima-Silva et al. (2009), which describes the effect of carbohydrate (CHO) availability on heart rate (HR) variability (HRV) during moderate and severe exercise intensities until exhaustion. While we congratulate the authors for having conducted a nice experiment, we wish to comment on the methodology used to examine HRV, to discuss the plausibility of their conclusions and also the interpretation that it is possible, using their measurements, to consider the mechanic versus autonomic determinants of HRV during exercise. Also the practical applications of the present findings remain obscure to us.First, it is worth mentioning that the interactions between the pulmonary and the autonomic nervous systems, known as the respiratory sinus arrhythmia phenomenon (Hayano et al. 1996), and their resultant impact on HRV indices, are exercise intensity-dependent. Second, exercise-induced changes in sympathetic-and vagal-related HRV indices display different time courses. And third, exercise intensity does not affect absolute [in power spectral density (PSD, ms 2 ) or in ms] and normalized [accounting for the total PSD or variance, in normalized unit (n.u.) or %] HRV indices in a similar fashion. Several authors [e.g. (Blain et al. 2005b;Cottin et al. 2006;Buchheit et al. 2007;Cottin et al. 2007)] have depicted two objective 'tipping points' [i.e., the first (VT1) and second (VT2) ventilatory threshold] which clearly demarcate three distinct exercise-intensity domains (i.e., below VT1, between VT1 and VT2 and above VT2, Fig.