Abstract There is emerging evidence linking fruit and vegetable consumption and cognitive function. However, studies focusing on the nutrients underlying this relationship are lacking. We aim to examine the association between plasma nutrients and cognition in a population at risk for cognitive decline with a suboptimal diet. The Mediterranean-DASH Intervention for Neurodegenerative Delay (MIND) trial is a randomized controlled intervention that examines the effects of the MIND diet to prevent cognitive decline. The primary outcome is global cognition. A multivariate linear model was used to investigate the association between blood nutrients and global and/or domain-specific cognition. The model was adjusted for age, sex, education, study site, smoking status, cognitive activities and physical activities. High plasma α-carotene was associated with better global cognition. Participants in the highest tertile of plasma α-carotene had a higher global cognition z score of 0⋅17 when compared with individuals in the lowest tertile (P 0⋅002). Circulating α-carotene levels were also associated with higher semantic memory scores (P for trend 0⋅007). Lutein and zeaxanthin (combined) was positively associated with higher semantic memory scores (P for trend 0⋅009). Our study demonstrated that higher α-carotene levels in blood were associated with higher global cognition scores in a US population at risk for cognitive decline. The higher α-carotene levels in blood reflected greater intakes of fruits, other types of vegetables and lesser intakes of butter and margarine and meat. The higher circulating levels of lutein plus zeaxanthin reflected a dietary pattern with high intakes of fruits, green leafy, other vegetables and cheese, and low consumption of fried foods. Objective nutrient markers in the blood can better characterize dietary intake, which may facilitate the implementation of a tailored dietary intervention for the prevention of cognitive decline.