Objective: To identify predictors of dietary change to and maintenance of a low-fat eating pattern (<20% energy from fat, ≥5 servings fruits/vegetables daily, and ≥6 servings grains daily) among a cohort of postmenopausal women. Candidate predictors included intrapersonal, interpersonal, intervention program characteristics, and clinical center. Design: Longitudinal study within the Women's Health Initiative Dietary Modification Trial. Dietary change was evaluated after 1 year of participation in the Women's Health Initiative Dietary Modification Trial, and dietary maintenance after 3 years. Subjects: Postmenopausal women aged 50 to 79 years at baseline who were randomized to the intervention arm of the Women's Health Initiative Dietary Modification Trial (n=19,541). Statistical analysis: Univariate and multivariate linear regression analysis was performed and associations evaluated between candidate predictors and each of the three dietary goals: percent energy from fat, fruit/vegetable servings, and grain servings. Results: Year 1 (change) predictors of percent energy from fat (P<0.005) included being younger (β=2.12; 70 to 79 years vs 50 to 59 years), more educated (β=-.69; college vs high school), more optimistic (β=-.07), attending more sessions (β=-.69), and submitting more self-monitoring records (β=-.74). At year 3 (maintenance), the predictors of percent energy from fat (P<0.005) included attending more sessions (β=-.65) and submitting more self-monitoring scores (β=-.71). The analytic model predicted 22% of the variance in fat intake at year 1 and 27% at year 3 (P<0.01). Conclusions: The strongest predictors of dietary change and maintenance were attending intervention sessions and self-monitoring dietary intake. Novel was the finding that optimism predicted dietary change.