Compressional sound velocities of shelf sediments in the South Sea of Korea, were measured in situ and in the laboratory for six cores. In situ sound velocity was measured using the Acoustic Lance (frequency of 7.5-15 kHz), while laboratory velocity was measured by the pulse transmission technique (frequency of 1MHz). Physical properties were relatively uniform with sediment depth, suggesting little effect of sediment compaction and/or consolidation. Average in situ velocity at each core site ranged from 1,457 to 1,488 m/s, which was less than the laboratory velocity of 1,503 and 1,604m/s. In muddy sediments the laboratory velocity was 39-47 m/s higher than in situ velocity. In sandy sediments, the difference was greater by an average of 116 m/s. Although the velocity data were corrected by the velocity ratio method based on bottom water temperature, the laboratory velocity was still higher than the in situ velocity (11-21 m/s in muddy sediments and 91 m/s in sandy sediments). This discrepancy may be caused by sediment disturbance during core collection and/or by the pressure of Acoustic Lance insertion, but it was most likely due to the frequency difference between in situ and laboratory measurement systems. Thus, when correcting laboratory velocity to in situ velocity, it is important to consider both temperature and frequency.