The global growing rates of cognitive decline and dementia, together with the absence of curative therapies for these conditions, support the interest in researching potential primary prevention interventions, with particular focus on dietary habits. The aim was to assess the association between polyphenol intake and 6-year change in cognitive function in the 'Seguimiento Universidad de Navarra' (SUN) Project, a Spanish prospective cohort study. Changes (final - initial) in cognitive function were evaluated in a subsample of 806 participants (mean age 66 (sd 5) years, 69·7 % male) of the SUN Project using the validated Spanish Telephone Interview for Cognitive Status-modified score. Polyphenol intake was derived from a validated semi-quantitative FFQ and matching food composition data from the Phenol-Explorer database. Multivariable linear regression models were used to evaluate the association between total polyphenol intake, polyphenol subclasses and cognitive changes. No significant association between total polyphenol intake and changes in cognitive function was found. However, a higher intake of lignans (ßQuintile (Q) 5 v. Q1 0·81; 95 % CI 0·12, 1·51; Ptrend = 0·020) and stilbenes (ßQ5 v. Q1 0·82; 95 % CI 0·15, 1·49; Ptrend = 0·028) was associated with more favourable changes in cognitive function over time, particularly with respect to immediate memory and language domains. Olive oil and nuts were the major sources of variability in lignan intake, and wine in stilbene intake. The results suggest that lignan and stilbene intake was associated with improvements in cognitive function.