Hydroacoustic techniques are becoming widely used in fishery resource assessment. There are many sources of variability and uncertainty in these techniques, and their accuracy and precision has been questioned. Studies at the University of Washington have evaluated the performance of hydroacoustic techniques in a variety of applications where substantial ground truth data were available. These studies have documented the magnitude of several variance components and shown that many of the larger sources of variability are biological rather than physical. [Work supported by Sea Grant.]