During the past decade, compositional analysis (CA) has been used widely in animal–habitat and resource selection studies. Despite this popularity, CA has not been tested for potential systematic biases such as incorrect identification of preferred resources. We used computer-simulated data based on known habitat use and availability parameters to assess the potential for CA to incorrectly identify preferred habitat use. We consider in particular the situation when available habitat categories not used by all animals are included in the resource selection analysis, with substitution of a relatively small value, such as 0.01, for each 0% utilization value. Progressively larger misclassification-error rates in preferred habitat use resulted from substituting progressively smaller positive values for each 0% utilization of a habitat category.