The effects of trampling are usually confounded by the diffuse impacts of the urbanization of sandy beaches. We performed a controlled experiment on a beach with low visitation rates to test the hypothesis that ghost crabs avoid building their burrows on impacted plots as a result of the compacted sediment, and they migrate to non-trampled areas. The sampling design encompassed 11 survey quadrats (6 × 6 m) above the strandline, including five trampled plots (100, 300, 900, 1500 and 3000 steps) and six non-trampled plots. The plots were sampled before and after 24, 48 and 72 h of experimental trampling. We found that the ghost crabs avoided building their burrows in only the 1500× and 3000× trampled plots after 24 h, but the avoidance was not related to sediment compactness. Additionally, the emersion time and escape distance from humans were significantly delayed in the most trampled plots, suggesting a lower surface activity and an avoidance of irregular (i.e. high micro-relief) sediment surfaces by ghost crabs, which might reduce their ability to perceive potential predators.