Eradicating feral pigs from island ecosystems can assist in restoring damaged biodiversity values and protect commercial industries such as agriculture. Although many feral pig eradications have been attempted, management decisions are often led by practitioner experience rather than empirical evidence. Few interventions have been guided by population models to identify harvest rates necessary to achieve eradication within a specified time frame, nor have they applied data on control effort and cost to evaluate the relative cost-effectiveness of proposed control strategies. We used effort and cost data from a feral pig-control program on Kangaroo Island, South Australia over 17 months to derive functional-response relationships between control effort (hours pig-1) and pig abundance for four control methods: (i) ground-based shooting, (ii) trapping with remote triggers, (iii) poison baiting, and (iv) thermal-assisted aerial culling. We developed a stochastic Leslie matrix with compensatory density feedback on survival and fertility to project population trajectories from an initial population (N0) of 250 female pigs with an estimated island-wide carrying capacity (K) of 2500 over 3 and 10 years for populations subjected to an annual harvest of 35% to 95%. We built functional-response models to calculate annual effort and cost for six cull scenarios across all harvest rates. We derived total cost and effort over 3- and 10-year projections from the sum of annual cost and effort within the projection intervals. Pig populations were reduced to < 10%N0based on harvest rates > 70% and 50% for culls of 3- and 10-year duration, respectively. In all scenarios except ′trapping only′, the total cost to reduce population to ≤ 10% ofN0decreased with increasing harvest proportion, with lower total costs incurred over 3 years compared to 10 years. The simulations suggest that the most cost-effective approach for most scenarios is to maximise annual harvest and complete eradication effort over the shortest periods.