Resting state electroencephalographic (EEG) activity in schizophrenia (SZ) is frequently characterized by increased power at slow frequencies, and/or a reduction of peak alpha frequency. Here we investigated the nature of these effects. As most studies to date have been limited by reliance on a priori frequency bands which impose an assumed structure on the data, we performed a data-driven analysis of resting EEG recorded in SZ patients and healthy controls (HC). The sample consisted of 39 chronic SZ and 36 matched HC. The EEG was recorded with a dense electrode array. Power spectral densities were decomposed via Varimax-rotated principal component analysis (PCA) over all participants and for each group separately. Spectral PCA was repeated at the cortical level on cortical current source density computed from standardized low resolution brain electromagnetic tomography. There was a trend for power in the theta/alpha range to be increased in SZ compared to HC, and peak alpha frequency was significantly reduced in SZ. PCA revealed that this frequency shift was due to the presence of a spectral component in the theta/alpha range (6-9 Hz) that was unique to SZ. The source distribution of this component involved mainly prefrontal and parahippocampal areas. Abnormal low frequency resting EEG activity in SZ was accounted for by a unique theta/alpha oscillation. Other reports have described a similar phenomenon suggesting that the neural circuits oscillating in this range are relevant to SZ pathophysiology.