Abstract: Acoustic methods could play an important role in monitoring vertical and horizontal fish distribution; including fish stock assessment, however the fish as acoustic target is not precisely specified still. The paper describes a new type of approach to express diel variability of acoustic parameter s_v, characterizing fish volume density. Multi-depth-channel integration of fish echoes was applied to estimate the average diel models of s_v dependence on time of a day. The final characteristics was calculated by computer as a vectoral field, showing quasi-continuous diel transformation of depth distribution of s_v values. Example based on data collected during October 1996 RV "Baltica" research cruise was given and discussed. Strong variability of s_v values with time and depth was exposed and clearly different diel phases of fish behaviour were separated. Results indicate a necessity of further research on fish target strength within separate periods of a diel cycle, to define more accurately biomass conversion factors. The method shown could enreach significantly systematic physiological studies on fish behaviour also.