Visiting or viewing nature environments can have restorative psychological effects, while exposure to the built environment typically has less positive effects. A classic view is that this difference in restorative potential of nature and built environments depends on differences in the intrinsic characteristics of the stimuli. In addition, an evolutionary account is often assumed whereby restoration is believed to be a hardwired response to nature's stimulus-features. Here, we propose the novel hypothesis that the restorative effects of a stimulus do not entirely depend on the stimulus-features per se, but also on the meaning that people assign to the stimulus. Participants conducted cognitively demanding tests prior to and after a brief pause. During the pause, the participants were exposed to an ambiguous sound consisting of pink noise with white noise interspersed. Participants in the "nature sound-source condition" were told that the sound originated from a nature scene with a waterfall; participants in the "industrial sound-source condition" were told that the sound originated from an industrial environment with machinery; and participants in the "control condition" were told nothing about the sound origin. Self-reported mental exhaustion showed that participants in the nature sound-source condition were more psychologically restored after the pause than participants in the industrial sound-source condition. One potential interpretation of the results is that restoration from nature experiences depends on learned, positive associations with nature; not only on hardwired responses shaped by evolution.