OBJECTIVE: To determine whether there are differences among measures of heart rate variability (HRV; traditional and nonlinear) after anesthesia and cardiac surgery. DESIGN: Prospective. SETTING: University hospital. PARTICIPANTS: Patients scheduled for cardiac surgery. INTERVENTIONS: None. Medical management was not varied as part of this study. MEASUREMENTS AND MAIN RESULTS: HRV was measured in 13 patients from electrocardiograms (ECGs) recorded before anesthesia, during anesthesia but before cardiac surgery, and on the first postoperative day. Anesthesia was induced with moderate-dose fentanyl. For each ECG, HRV was measured from series of 400 heartbeat intervals using standard deviation (SD), approximate entropy (ApEn), and point correlation dimension (PD2). Multivariate repeated-measures analyses of variance on ranks and Spearman correlations were performed. All HRV measures decreased significantly with anesthesia. Postoperatively, ApEn recovered to original values. PD2 and SD did not recover with consciousness and were significantly less than original values. Correlations among ApEn, PD2, and SD were weak. CONCLUSIONS: Nonlinear measures of HRV differ among themselves after anesthesia and cardiac surgery. The use of multiple nonlinear and traditional measures may improve the effectiveness of using HRV to assess the cardiovascular system.