The relation between mid-water trawl catchability and its construction, trawling regime, krill aggregations distribution in the study area is discussed. It is shown that while the trawl selectivity affects krill length composition in each trawl sample, the trawl catchability affects the estimation of total krill lengths distribution, obtained on the basis of trawl samples from different catched in the study area. Several approaches to trawl samples processing are discussed. It is shown that the total distribution of krill length, obtained using weighting of trawl samples (usually 200 ind.) against the catch, may considerably differ (statistically significantly) from that calculated on the basis of trawl samples weighting against the catch corrected taking in account the trawl catchability for each trawling. The example of trawl samples processing for the study area off Elephant Island is presented. Difference of total krill lengths distribution, calculated using different methods, results in respective difference of density estimates by 48%.