First attempts were made in the early 1980’s to estimate the krill and pelagic food consumption by Antarctic demersal fish. These estimates were extended to the mesopelagic realm and the high-Antarctic Zone in the late 1980’s and early 1990’s when these areas were exploited commercially and a larger number of food studies were conducted. Currently, the best estimates of krill consumption by fish are 23 – 29 . 106 tonnes of krill and other pelagic prey taken annually by demersal fish and 7 – 44 . 106 tonnes taken by mesopelagic fish in the Atlantic Ocean sector only. However, these estimates still have wide confidence limits. Major shortcomings of the consumption estimated for mesopelagic fish are the validity of hydro-acoustic biomass estimates conducted in the late 1980’s and the lack of quantitative food consumption data in some abundant myctophid species. Major shortcomings of the consumption estimates of demersal fish are the inaccurracy of biomass estimates for some species, the shortness of most food studies which do not adequately grasp the opportunistic feeding habits of many demersal fish and the scarcity of quantitative winter food studies. However, it is evident from this review that the importance of krill in fish diets varies with time and location, and with the suite of prey types available in different regions in the Southern Ocean. Models of food web and ecosystem dynamics need to account for the role important taxa, such as myctophids, Champsocephalus gunnari, and other abundant channichthyids and nototheniids play as predators of krill and other pelagic resources. Furthermore, the effects of large changes in abundance and community structure of fishes brought about by industrial fishing needs to be considered.