We considered the problem of conserving small populations of prey by reviewing observational and field experiments in search of general answers to the question of how predators impact upon their prey. We then explored the likely effectiveness of predator control as a conservation tool by simulating a small, fragmented population of prey. We chose red grouse in Britain as an example of this general predicament. An individual-based population simulation model was used to simulate the fate of a small red grouse population. The model was used to investigate the relative impacts of predation, habitat changes and stochastic factors associated with small population size on the viability of the population. Although most predator–prey studies have been undertaken in temperate regions and only four studies found evidence of regulation, some generalizations arose. Predation can limit populations but generally did not drive the prey population to extinction. Extinction was more likely in cases involving an introduced predator on islands or in isolated prey populations. The simulation model showed that at small population sizes the stochastic factors were always significant, whereas environmental variation and habitat change had the greatest impact on large populations over the long term; predation was only a significant threatening factor when it occurred at a high rate, was more severe over the long term when predators took a constant proportion of the prey base, and when predation was across all age classes.