A sonic pulse system, similar to that used at Lamont Geological Observatory for seismic model experiments, was used aboard the Research Vessel VEMA during the summer of 1954 to determine high frequency seismic velocities in fresh deep sea sediment cores. Velocity profiles were obtained from 26 cores covering a wide range of lithologies and ages (Recent to Miocene). Density, porosity, median grain size, sorting, carbonate content, and salt content were also measured. The compressional wave velocity in the ocean‐bottom unconsolidated sediments studied is well represented by the equation: [Formula: see text] where v′=compressional wave velocity in km/sec ϕ=median grain size in phi units γ=percentage of HCl soluble material η=porosity. Many measurements gave velocities less than the velocity of sound in sea water. Most of the low carbonate samples followed a velocity‐porosity relation given by the Wood (1941) equation. The regression coefficient, −.44η, agrees well with the average slope of the Wood equation over the observed porosity range. High carbonate and large median grain size samples gave velocities above that predicted by the Wood equation. These higher velocities are explained as the combined result of shear strength and low effective porosity in the samples. The highest velocities were found in slowly deposited sediments. The degrees of sorting of the sediments had no observable effect on the seismic velocities except that unexplained variations were greater for more poorly sorted materials. No correlation between velocity and age was evident in the sediments studied. The effect of temperature, between 40 and 80°F. on compressional velocity in sediments may be explained by changes in elastic properties of the water fraction alone. The effect of compaction in the upper 15 or 20 feet of homogeneous sediments produced a change in seismic velocity not greater than 1 or 2 percent. Attenuation was greater in the coarse‐grained high‐velocity sediments than in sediments of smaller grain size.