Recreation negatively affects wildlife by influencing animal behaviour vital to reproduction and survival. Such non-consumptive effects of perceived predation risk are mainly studied in ground-breeding birds. However, if anti-predator responses characterize bird species generally, so should non-consumptive effects of perceived predation associated with human recreation. Moreover, as individuals consistently differ in behaviours linked to anti-predator responses, they should also differ in responses to recreation, with bolder birds being less affected. To test this key prediction, we quantified effects of human recreation pressure on a cavity-breeding passerine. We uniquely quantified human recreation pressure over a substantial (8-year) period within twelve nest-box populations of the great tit Parus major, assayed annually for reproductive parameters. We detected considerable spatial variation in recreation pressure. In plots with high recreation pressure, we found strong support for birds breeding further away from highly frequented paths and birds producing smaller clutches; we also found moderate support for birds producing fewer fledglings. These detrimental effects did not vary with behavioural proxies of an individual’s risk-taking phenotype (exploratory activity). This implies that effects of recreation pressure apply to the average bird, and extend to species (like forest birds) not previously considered.