Running shoe cushioning has become a standard method for managing impact loading and consequent injuries due to running. However, despite decades of shoe technology developments and the fact that shoes have become increasingly cushioned, aimed to ease the impact on runners’ legs, running injuries have not decreased. To better understand the shoe cushioning paradox, we examined impact loading and the spring-like mechanics of running in a conventional control running shoe and a highly cushioned maximalist shoe at two training speeds, 10 and 14.5 km/h. We found that highly cushioned maximalist shoes alter spring-like running mechanics and amplify rather than attenuate impact loading. This surprising outcome was more pronounced at fast running speed (14.5 km/h), where ground reaction force impact peak and loading rate were 10.7% and 12.3% greater, respectively, in the maximalist shoe compared to the conventional shoe, whereas only a slightly higher impact peak (6.4%) was found at the 10 km/h speed with the maximalist shoe. We attribute the greater impact loading with the maximalist shoes to stiffer leg during landing compared to that of running with the conventional shoes. These discoveries may explain why shoes with more cushioning do not protect against impact-related running injuries.