Diet is an important modifiable lifestyle factor related to dementia risk. Yet, the role of midlife dietary changes is unclear. The goal is to investigate whether midlife healthy dietary changes are associated with late-life dementia risk. Data were collected within the Cardiovascular Risk Factors, Aging, and Dementia (CAIDE) population-based cohort study (n = 2000) (mean baseline age = 56 years). Participants returned for two late-life re-examinations (mean age = 70 and 78 years). Self-reported midlife diet was measured in a sub-sample (n = 341) (mean total follow-up = 16.8 years). Changes in specific dietary components (fats, vegetables, sugar, salt) were measured in midlife. Dementia diagnoses were ascertained with detailed examinations. Analyses adjusted for potential confounders. Total midlife healthy dietary changes (improving quality of fats, increasing vegetables, decreasing sugar and salt) were associated with a reduced risk of dementia (fully adjusted odds ratio (OR) 0.41, 95% confidence interval (CI) = 0.20–0.85). In contrast, when each factor was assessed individually, associations were not significant. This study is the first to show that beneficial midlife dietary changes are associated with a reduced dementia risk later in life. The results highlight the importance of targeting dietary patterns, where various food items may have synergistic effects.