Estimates of fish abundance based on arrival rates and numbers present at baited cameras allow multiple, replicate assessments where data are not available by other means. Unfortunately such estimates are strongly affected by the assumed behaviour of the fish species concerned. Three of the possible foraging strategies of deep-sea fish were modelled and the likely patterns of fish arrival calculated for the same fish density, swimming and current velocities and odour plume properties. Cross-current foraging resulted in the highest numbers of fish at bait, with arrival rates that fitted well to field data. The sit-and-wait strategy produced lower arrival rates with passive drifting animals arriving slowest. Each model produces a distinctive pattern of animal arrivals that may be diagnostic of each foraging strategy. The advantages, disadvantages and likely metabolic and sensory demands of each strategy are discussed.