OBJECTIVES: Debate continues regarding the stability of preferences across time and illness. The purpose therefore was to determine if the experience of depressive disorder changes patients' risk attitudes and elasticity for time, components of their standard gamble (SG) and time trade-off (TTO) utilities. This study tests the explanatory power of the Medical Care System Access Framework for depression experience impact on patient preferences while controlling for patient population characteristics. METHODS: This study used two years of data from Partners in Care, a group-level randomized controlled trial of quality improvement programs for depression. For 1218 primary care patients with depression, we examined single-item SG and TTO utilities at baseline and 24 months. Logistic regressions identified factors associated with patients' willingness to take risks and trade time and examined trends in utilities of individuals with and without remission of depression as measured by the Center for Epidemiologic Studies Depression screener and the World Health Organization Composite International Diagnostic Instrument. RESULTS: A dose-response gradient indicating greater willingness to take risks or trade time was found as depression increased. Patients who continued to be depressed at 24 months were nearly 3 times more likely to be willing to assume risk or trade time than patients whose depression remitted. Willingness to assume risk or trade time increased in patients who continued to be depressed at 24 months. However the SG appeared to lack sensitivity in patients whose depression remitted. Remission of depression was associated with a decrease in willingness to trade time as expected and a paradoxical increase in willingness to assume risk. CONCLUSIONS: Patients with depression appear to use a single-item SG measure differently at baseline and 24 months. This could be due to changes in risk attitude resulting from health experiences. Changes in values may confound the use of single-item SG utilities as measures of outcomes.