The AMLR hydroacoustic system used for surveys in the vicinity of Elephant Island and King George Island during 1987-1989 has hardware and software components. The method of analysis proceeds from the basic integration data (lm bins by 1min outputs) to the stratification of these data into blocks of area and finally a total estimate of abundance and statistical confidence limits about the block estimates and the total biomass in the surveyed area. The process depends on: 1) a calibrated, towed hydroacoustic system; 2) real-time analysis of data; and 3) using a structured database which ensures data integrity and repeatability of analytic results. The variability inherent in the distribution of krill mandates the use of covariance methods for the calculation of confidence intervals. This often requires post-cruise stratification of the survey data into statistical sub-areas. The intermediate and final results of these analyses may also be used to produce other data products for comparison with environmental or other biological data.