As self-reports, measures of patient utility are susceptible to the effects of cognitive biases in patients. This article presents often overlooked problems in these measures by outlining cognitive processes involved in patient self-report. It is argued that these measures: 1) require overly complex mental operations; 2) fail to elicit thoughtful response by default; 3) may be biased by patients' mood; 4) are affected by both researchers' choice of measurement instruments and patients' choice of judgment strategies; 5) tend to reflect the disproportionate influence of patients' values that happen to be recallable at the time of measurement; and 6) are affected by patients' fear of regret. It is suggested that solutions for these problems should involve: a) improving the methods of administration; b) developing measures that are less taxing to patients; and c) redefining the concept of patient utility as judged, as opposed to retrieved, evaluation. Published by 2000 Elsevier Science Inc.