Previous tests of ideal free distribution (IFD) under continuous input conditions have demonstrated that more profitable patches tend to be relatively underused compared to that predicted by the theory. We tested the hypothesis that competitors' perceptual constraints of resource distribution cause this deviation from the IFD. A laboratory experiment was conducted to determine whether additional information on food distribution by a light cue that indicates the food input point improves the IFD theory's fit to the distribution of clone red-spotted masu salmons (Salmonids), Oncorhynchus masou ishikawai, that had been conditioned to the light as a cue indicating the site with a higher input rate. In the treatments without a light cue, the distribution of fish was closer to a random pattern than an IFD. In contrast, in the treatments with light cue, the distribution of fish was closer to the expected value of an IFD rather than to a random pattern, supporting the perception-limit hypothesis. The distribution and the pattern of resource use by fish in the treatments without the light cue were best explained by the perception-limit model. Our results suggest that it is perceptual constraints that cause deviation from the IFD.