The lower boundary of the methane hydrate stability zone in continental margin sedi-mentsis often marked by a strong, phase reversed reflection subparallel to the seafloor,called the bottom simulating reflector (BSR). High resolution multichannel seismic(MCS) data from the Yaquina Basin offshore Peru at 8 deg S show a BSR that is vary-inglaterally in amplitude as well as in continuity. The amplitudes of the reflectionsabove the BSR also vary with the appearance of the BSR. Where the BSR is strong,the reflections above it are weaker compared to areas where the BSR is weak. Andalthough the strong part of the BSR is underlain immediately by strong reflections,reflections several hundred meters beneath the BSR appear weaker than those wherethe BSR is weak. This variation indicates significant heterogeneity in the distribu-tionof gas and gas hydrate in this area. Chemoherms observed at the Yaquina Basinsea floor indicate the presence of free gas in the sediments up to the seafloor. Thepresence of gas and gas hydrate within the sediment sequence significantly influencesthe P-wave velocity in the affected layers. Therefore a detailed analysis of velocityvariations enables to understand the apparently different conditions for the formationof gas hydrate along the BSR and the migration paths of the free gas. Ocean bot-tomseismometer (OBS) data from profiles coincident with the MCS data can providesuch detailed velocity depth information. Velocity analysis from OBS data included2D-ray tracing and 1D-interval-velocity analysis by means of DIX-inversion. In orderto find a trade-off between vertical resolution and minimization of errors caused bythe sensitivity of the DIX formula to velocity variations in thin layers, the data haveundergone a Kirchhoff wave-equation datuming and adjacent coherence filtering wasapplied to the data to eliminate the one sided travel path through the water columnof the OBS-observations. The derived velocity structure confirms the interpretation ofthe reflection pattern in terms of gas and gas hydrate distribution.