Hypothetical bias is a persistent problem in stated preference studies. We propose and test a method for reducing hypothetical bias based on the cognitive dissonance literature in social psychology. A central element of this literature is that people prefer not to take inconsistent stands and will change their attitudes and behaviour to make them consistent. We find that participants in a stated preference willingness-to-pay study, when told that a nonhypothetical study of similar goods would follow, state significantly lower willingness to pay (WTP) than participants not so informed. In other words, participants adjust their stated WTP to avoid cognitive dissonance, that is, taking inconsistent stands on their WTP for the good being offered.