Objective-We examined the association of treatment preferences with treatment initiation, adherence, and clinical outcome among depressed mid-life and elderly primary care patients.Methods-60 primary care participants meeting DSM-IV criteria for major depression were randomized to receive treatment congruent or incongruent with their primary stated preference. Participants received either 20 weeks of escitalopram as monitored by a care manager, or 12 weekly sessions of interpersonal psychotherapy followed by 2 monthly booster sessions. Adherence to treatment and depression severity were reassessed at weeks 4, 8, 12, and 24.Results-Participants expressed stronger preferences for psychotherapy than antidepressant medication. Preference strength was a more sensitive measure of outcomes than congruence versus incongruence of preference with the assigned treatment. Across age groups, preference strength was significantly associated with treatment initiation and 12-week adherence rate, but not with depression severity or remission.Conclusions-A continuous measure of preference strength may be a more useful measure in clinical practice than preferences per se. Future research should focus on whether and how greater facilitation of the patient-clinician treatment decision-making process influences clinical outcome.Treatments of depression in primary care settings are effective yet most depressed adults (1), particularly older ones (2), do not receive appropriate care. Even when guideline-based treatments are provided, patients often do not fully participate in them. Not surprisingly, therefore, randomized clinical trials have reported substantially poorer outcomes for "intent to treat" than "treatment completer" cohorts (3), indicating a need for strategies that maximize treatment participation.A patient's decision not to initiate or complete treatment may stem from disappointment or dissatisfaction with the treatment offered by the clinician. While medications are the predominant intervention offered depressed primary care patients, 50%-86% of them prefer a psychosocial intervention (4-7). Thus, many patients conceivably refuse treatment offered in primary care because psychotherapy is not an available option.In psychiatric outpatient settings, treatment preferences have been addressed through "negotiated treatment plans" whereby clinicians elicit patient requests and encourage their participation in treatment planning. Patient reports of greater participation in such negotiations have been associated with greater levels of satisfaction, sense of feeling helped, and adherence to treatment plans (8,9). Studies of mid-life patients in the primary care sector have endorsed the value of a negotiated treatment plan and the importance of patients playing active rather than passive roles in formulating it. Such participation enhances the patient's likelihood of initiating treatment and his/her satisfaction with it (10-12).Despite these benefits, the few studies examining treatment negotiation and clinical outcome have...