Abstract: OBN data have better quality than streamer data, but OBN acquisition is complicated and expensive. Thus, it is necessary to perform a detailed assessment of OBN and streamer data from the perspectives of imaging and structure and reservoir characterization as well as OBN comparative advantages for hydrocarbon exploration and development. The case study deals with deep-water sub-salt formations. We first examine the discrepancies between OBN and streamer data in offset, azimuth, illumination, signal-to-noise ratio, and seismic velocity field in accordance with their different geometries. In view of the challenges in complicated sub-salt formation imaging, the comprehensive analysis involves diffracted-wave imaging, fault imaging, basement imaging, and vertical and lateral resolutions of these two data types and their interpretation results of structure depth and inversion. With respect to abundant low frequencies in OBN data, we use filtered log data to investigate their contribution to the characterization of geologic boundaries and propose a method of resolution calculation using statistical wavelet. Compared with streamer data, OBN data show better quality owing to large-offset wide-azimuth acquisition and consequent better illumination and more accurate velocity field. Similar sub-salt structural features and trends of inversion results could be obtained using streamer and OBN data, but the latter performs better in detailed structure and reservoir characterization. In summary, OBN data feature better imaging, more distinct fault and basement reflections, higher resolution, wider band, more abundant low frequencies, smaller error of structure depth, and more detailed inversion results compared with streamer data. Research results may be used for reference in the study of streamer and OBN data in other oilfields.