Some fishes and sea turtles are distinct from ectotherms by having elevated core body temperatures and metabolic rates. Quantifying the energetics and activity of the regionally endothermic species will help us understand how a fundamental biophysical process (i.e. temperature-dependent metabolism) shapes animal ecology; however, such information is limited owing to difficulties in studying these large, highly active animals. White sharks, Carcharodon carcharias, are the largest fish with regional endothermy, and potentially among the most energy-demanding fishes. Here, we deployed multi-sensor loggers on eight white sharks aggregating near colonies of long-nosed fur seals, Arctocephalus forsteri, off the Neptune Islands, Australia. Simultaneous measurements of depth, swim speed (a proxy for swimming metabolic rate) and body acceleration (indicating when sharks exhibited energy-efficient gliding behaviour) revealed their fine-scale swimming behaviour and allowed us to estimate their energy expenditure. Sharks repeatedly dived (mean swimming depth, 29 m) and swam at the surface between deep dives (maximum depth, 108 m). Modal swim speeds (0.80-1.35 m s −1 ) were slower than the estimated speeds that minimize cost of transport (1.3-1.9 m s −1 ), a pattern analogous to a 'sit-and-wait' strategy for a perpetually swimming species. All but one shark employed unpowered gliding during descents, rendering deep (>50 m) dives 29% less costly than surface swimming, which may incur additional wave drag. We suggest that these behavioural strategies may help sharks to maximize net energy gains by reducing swimming cost while increasing encounter rates with fast-swimming seals.