The extent to which variation in posture and respiration can confound pre-ejection period and respiratory sinus arrhythmia (RSA) as indices of cardiac sympatho-vagal activity was examined. Within-subjects changes in these measures were assessed in 36 subjects during different postures and (paced) respiratory frequencies. Changes from supine to sitting to standing led to reduced RSA values and longer pre-ejection periods, reflecting the known decrease in vagal but not the increase of sympathetic activity. Multilevel path analysis showed that within-subjects changes in sympatho-vagal balance were faithfully reflected by changes in interbeat interval, but imperfectly by changes in RSA and pre-ejection period. It was concluded that pre-ejection period should be stratified for posture and RSA for respiratory frequency to reliably index changes in sympatho-vagal balance when these factors are prone to change (e.g., during 24-h ambulatory recording). Copyright © 2005 Society for Psychophysiological Research.