Research has shown that hedonic-contrast effects are a ubiquitous and important phenomenon. In eight studies (N = 4,999) and four supplemental studies (N = 1,809), we found that hedonic-contrast effects were stronger for negative outcomes than for positive outcomes. This asymmetric-contrast effect held for both anticipated and experienced affect. The effect makes risks that include gains and losses more attractive in the presence of high reference points because contrast diminishes the hedonic impact of losses more than gains. We demonstrated that the effect occurs because people are generally more attentive to reference points when evaluating negative outcomes, so drawing attention to reference points eliminates the asymmetric-contrast effect.