Extant theories of decoy effects on evaluations of attribute values were assessed with respect to their ability to account for a one-dimensional analogue of the asymmetric dominance effect. Parducci’s (1965, 1995) range-frequency theory, Krumhansl’s (1978) distance-density model, Tversky’s (1977) diagnosticity principle, and reference point theories (e.g., Holyoak & Mah, 1982) were unable to account for this effect. One version of Helson’s (1964) adaptation-level theory and our comparisoninduced distortion theory (Choplin & Hummel, 2002) were able to account for the qualitative effect. Quantitative fits revealed that comparison-induced distortion theory provides a better account of this effect than does adaptation-level theory. These results suggest that, in some cases, biases created by language-expressible magnitude comparisons mediate the effects of decoys on evaluation.