Comparative analysis of krill length distribution in the echosurvey area, obtained with trawl and acoustic methods is presented. Besides, the acoustic estimate of krill length composition was obtained on the basis of the empiric model by Greene et al. (1991) using estimates of target strength in situ at the frequency of 120 kHz as measured with echosounder EK-500 (split-beam) principle). Krill was caught with the commercial trawl RT 741448. It was shown that differences between krill length distribution predicted from the acoustic model by Greene et al. (1991) and those from trawl catches were considerable (statistically significant). Besides, krill length estimates, obtained with acoustic method, were significantly lower (20-25%) those obtained from catch analysis. Utilization of trawl and acoustic methods of krill length estimation in echosurveys implemented in practice as target strength (TS) in situ measurement with EK-500 and calculation of TS talc. on the basis of krill length composition in catches stipulates significant difference of krill density estimates obtained for the same population. In some cases such difference reached 200%. Significance of trawl samples in assessment of krill length composition during echosurveys was shown.