This study investigates the predatory behaviour of Cancer novaezelandiae (Decapoda : Brachyura : Cancridae), a subtidal crab from Lyttelton Harbour, Banks Peninsula. Cancer crabs were obtained at bimonthly intervals over a 22 month sampling period. Examination of stomach contents suggested that C.novaezelandiae was an opportunistic predator feeding on a wide variety of food items. Molluscs composed the major components of diet, with bivalves and gastropods comprising 35 % of all food items. Crustacea were the next most important group comprising 20 % of all food items. Amphipods, isopods and crabs were the most frequent crustaceans found. Other food types included fish, sponges, coelenterates and plant matter in relatively low abundances. No variation in dietary composition was evident with crab sex, size or season. The functional morphology of the chelae, mouthparts and gastric mill were examined for two size-classes of crab (60.0-70.0 and 120.0-130.0 mm carapace width). The mouthparts and gastric mill were generalised in structure and were typical of large, predatory brachyurans. Large crista dentata of the third maxilliped and sharp, rounded mandibles, together with large, heavily chitinised gastric mill ossicles suggested that C.novaezelandiae was well equipped to macerate course, particulate matter. No variation in either mouthparts and gastric mill structure was evident with crab size or sex. Left and right chela of C.novaezelandiae were morphologically similar in shape and dental pattern, and no difference was found between sexes. Relative growth of the chelae was described from measurements made on propodus height and length using carapace width as the reference dimension. Log-transformed regressions were used to test for allometric growth and lines were compared using t-tests. No significant difference was found in the growth of the propodus between left and right chelipeds, and for males and females. Relative growth of the propodus was positively allometric which remained constant throughout crab growth. No discontinuity in growth of the propodus was evident for male and female crabs (however for male crabs over 110 mm carapace width there was a slight indication of increased propodus growth). Both left and right chela of male and female crabs had a large diastema, and mean mechanical advantage of 0.367 and 0.375 respectively. The high mechanical advantage of the chela in C.novaezelandiae appeared to remain constant throughout crab growth, allowing a compressive force in excess of 496 kN.m⁻². The polyfunctional chela are therefore, capable of holding, manipulating and crushing a variety of prey shapes. Predator-prey experiments were undertaken in the laboratory using male crabs of three size classes (55.0-65.0, 80.0-90.0 and 105.0-115.0 mm carapace width). Prey used included blue mussel Mvtilus edulis aoteanus, cockle Chione stutchburyi, spotted whelk Cominella maculosa and catseye Turbo smaragdus. C.novaezelandiae adopted five distinct techniques to open bivalve and three techniques to open gastropod shells. The particular opening technique used was influenced by crab size, prey size and success of previous attempts. Small molluscs were usually opened by direct crushing by the chela and/or mouthparts. Splitting of the umbone valves and removal of the aperture lip were the most successful techniques used to open large bivalves and gastropods respectively. Handling times increased exponentially with prey size, with large crabs requiring less time to open prey than small crabs. Prey species influenced handling times with bivalves opened more quickly than gastropods. Critical maximum prey size increased with crab size, while minimum critical size was smaller for small crabs. Different sized crabs consumed similar numbers of prey, but energy intake increased with crab size. Profitability curves were derived and tested with regard to optimal prey size and species. When presented with a size range, crabs selected prey sizes that minimized time spent foraging. However when presented with different prey species, crabs maximized energy intake by selecting optimum prey species. Findings suggested C.novaezelandiae exhibited a flexible foraging behaviour, which allows crabs to maximize feeding efficiency. This was discussed in the context of the ecology of the crab and the Energy Maximization Premise (Elner and Hughes, 1978).