Description: Pinnipeds show huge inter-specific variability in the lengths of time that they spend on land and at sea over the year and consequently in the range over which they forage. The reasons for the inter- and intra-specific variability are not immediately obvious although each species-specific strategy presumably represents an idealized solution which integrates maximized energy acquisition at sea with optimized time spent at the haul-out site. Foraging behaviour, divided into transit time spent commuting between the haul-out site and foraging zone, as well as the transit between the water surface and the hunting depths once the animals are in the foraging zone, is liable to play a key role in this. This thesis examines the foraging behaviour of two pinniped species, the harbour seal Phoca vitulina and the Weddell seal Leptonychotes weddellii in order to tease out the precise factors that modulate the length and duration of foraging trips compared to haul-out periods. Methodological advances in technology are developed and tested to achieve this. Dead-reckoning principles contained within an archival tag were used to examine the fine-scale movement of animals underwater and, via consideration of the logged parameters, to determine how animals divided their time into various activities and how these were related to locality. The location of feeding behaviour was determined by using the dead-reckoner in conjunction with an inter-mandibular angle sensor (IMASEN). Consideration of time allocated to various phases of the dives conducted, together with the data on animal speed, pitch and roll indicated that harbour seals have specific travelling dives, prospecting dives, prey-searching dives and sleeping dives. Apart from the sleeping dives, these are also mirrored by Weddell seals. Track tortuosity, which is believed to correlate with prey searching behaviour, was highest in water deeper than 10 m and starting at distances of 15 km from the haul-out site for harbour seals. This concurred with seal density and allowed identification of the most important foraging sites for this species in the study area. Weddell seals foraged directly beneath their haul-out sites on the ice and therefore had no variable tortuosity in their tracks that could be allocated to horizontal travel between foraging and haul-out sites. Based on these parameters, the total time spent in a single foraging plus haul-out cycle could be subdivided into time allocated to active prey searching (which only occurs during the bottom phase of the dives) and 'other' time, which comprises time at the haul-out site plus time associated with travel between the foraging site and the haul-out site. A measure for foraging efficiency is proposed consisting of the total time allocated to searching for prey divided by the total time spent in one foraging plus haul-out cycle. Pinnipeds that are very efficient presumably need high prey densities at their foraging site so that the inverse of this efficiency gives a rough measure of a time-based 'prey- acquisition ration' (PAR). This index was modelled using data gained from the two seal species to examine how the extent of horizontal and vertical travel (representing proximity to prey and depth distribution of prey, respectively) modulates the PAR so that, ultimately, consideration of real prey density coupled with variation in PAR as a function of prey proximity determines where seals can feed. The trends illustrate why far-ranging pinnipeds need to have a large body mass and how this tends to make them highly proficient divers.