Strong bottom-simulating reflectors (BSR) have been mapped over a region of approximately 50,000 km2 on the southeastern U.S. margin and have been associated with possible abundance of natural gas hydrates. In June 1992, coincident single-channel seismic and wide-angle ocean bottom seismic data were acquired in the region, focusing on the Blake Ridge and the Carolina Rise. Wide-angle reflections from the BSRs were clearly observed at offsets up to ~6 km. Joint travel time inversion was conducted with wide-angle and vertical-incidence data in order to explore possible regional variation, and the resultant two-dimensional average velocity models imply higher background velocities on the Carolina Rise. Full waveform inversion was then performed to determine the seismic origin of the BSRs. The best fit model shows a similar low velocity (~1.4 km/s) beneath the BSR at both sites, indicating trapped free gas with low saturation (<10%). The inversion results also indicate that a thin, high-velocity wedge, with a maximum velocity of ~2.3 km/s, is present just above the Blake Ridge BSR. Sediment reflectivities were also calculated, and higher reflectivities are observed on the Carolina Rise. An increase in reflectivity below the BSR seems to correspond to the gas-bearing zone at both sites. Concentration of hydrates were estimated based on these velocity models. Whereas average hydrate concentration of 3% of the total sediment volume is suggested for the lower half of the hydrate stability zone at the Blake Ridge, only a very low average concentration of hydrate can be expected at the Carolina Rise. The hydrates seem to be concentrated near the base of the hydrate stability zone, and the maximum hydrate concentration is estimated as ~20% at the Blake Ridge and ~7% at the Carolina Rise, both of which are too high to be explained by in situ biogenic activity only and require some secondary accumulation mechanism. It is suggested that hydrate recycling caused by the stability field migration may have effectively condensed hydrates at both sites. Additional enhancement by upward fluid expulsion may also be viable for the Blake Ridge.¿ 1997 American Geophysical Union