Respiratory sinus arrhythmia (RSA)--the peak-to-peak variations in heart rate caused by respiration--can be used as a noninvasive measure of parasympathetic cardiac control. In the present study four strategies to increase RSA amplitude are investigated: (1) biofeedback of RSA amplitude, (2) biofeedback of RSA amplitude plus respiratory instructions, (3) respiratory biofeedback, and (4) respiratory instructions only. All four procedures produce a significant increase of RSA amplitude from the first physiological control trial compared to baseline. This increase is faster for the groups that received respiratory biofeedback and respiratory instructions only than for the two groups that received biofeedback of RSA amplitude, the increases being equivalent for the four groups in the third session. All subjects of the group that received biofeedback of RSA amplitude only reported respiratory strategies in order to achieve the increase in RSA. Possible clinical implications of these results for parasympathetic cardiac control and cardiovascular disorders are discussed.