Purpose: For depressed older primary care patients, this study aimed to examine (a) characteristics associated with depression treatment preferences; (b) predictors of receiving preferred treatment; and (c) whether receiving preferred treatment predicted satisfaction and depression outcomes. Design and Methods: Data are from 1,602 depressed older primary care patients who participated in a multisite, randomized clinical trial comparing usual care to collaborative care, which offered medication and counseling for up to 12 months. Baseline assessment included demographics, depression, health information, prior depression treatment, potential barriers, and treatment preferences (medication, counseling). At 12 months, services received, satisfaction, and depression outcomes were assessed. Results: More patients preferred counseling (57%) than medication (43%). Previous experience with a treatment type was the strongest predictor of preference. In addition, medication preference was predicted by male gender and diagnosis of major depression (vs dysthymia). The collaborative care model greatly improved access to preferred treatment, especially for counseling (74% vs 33% in usual care). Receipt of preferred treatment did not predict satisfaction or depression outcomes; these outcomes were most strongly impacted by treatment condition. Implications: Many depressed older primary care patients desire counseling, which is infrequently available in usual primary care. Discussion of treatment preferences should include an assessment of prior treatment experiences. A collaborative care model that increases collaboration between primary care and mental health professionals can increase access to preferred treatment. If preferred treatment is not available, collaborative care still results in good satisfaction and depression outcomes. (Contains 1 figure and 3 tables.)