Predators can have non-consumptive effects on their prey by causing anti-predator responses such as changes in behaviour. These effects may vary with the number of predators, which determines per capita predation risk. Predator density and cue concentration have been shown to affect prey responses in aquatic predator–prey systems; however, there are fewer tests in terrestrial systems. Here, we test the effects of predator density on prey dispersal and body growth in a system of predatory mites and their spider mite prey reared on leaf patches. Groups of prey were exposed to a low or high predator density level, or no predators. Prey dispersed in the presence of a predator and higher predator density led to greater prey dispersal. Growth in adult body size after maturation was reduced in the presence of a predator, and this effect was greater with a higher predator density, most likely related to a reduction in time spent feeding. Experiments were also conducted to test the effects of predator density mediated by predatory cues alone (previous presence of predators on the leaf patch). Spider mites were more likely to disperse when the patch had previously contained a higher density of predators; however, there was no effect of previous density level on body growth. These findings show that the non-consumptive effects of a predator on spider mites can depend on predator density and provide some evidence that chemical cues play a role in this density dependence. As these changes are likely to affect the predation rate and prey population growth rate, they are also likely to have consequences for the predator–prey dynamics.