Elevated rates of cardiac morbidity have been frequently reported in major depressive disorder (MDD) patients as a result of the relationship between autonomic dysfunctions and varied cardiovascular activity. Heart rate variability (HRV) analysis is an important and non-invasive way for assessing the variances in autonomic nervous system activity of MDD patients. In spectral domain, HRV analysis is usually done by either Fourier transformation (FT) or discrete wavelet transformation (DWT) to divide the data into lowfrequency (LF) and high-frequency (HF) bands. However, while FT is not a proper method for non-stationary HRV data, DWT does not exactly produce required frequency ranges of each LF and HF bands. The purpose of the present study is to investigate the spectral HRV measures obtained by wavelet packet transform (WPT) with absolutely excellent approximation to predefined frequency ranges of bands. Eighteen healthy controls and age- and gender-match eighteen patients with MDD were participated in this study. Sympathovagal balance (LF/HF ratio) that reflects the variation of sympathetic and parasympathetic activities was compared between two groups. Individuals with depression had a significantly higher LF/HF ratio. Our findings suggest that dysfunctions in coordination between sympathetic and parasympathetic nervous system activity in MDD patients can be evaluated by WPT based HRV analysis with high resolution decomposition for required LF and HF bands.