Aim It is well known that consumption of isoflavones reduces the risk of cardiovascular disease. However, the effectiveness of isoflavones in preventing dementia is controversial. A number of intervention studies have produced conflicting results. One possible reason is that the ability to produce equol, a metabolite of a soy isoflavone, differs greatly in individuals. In addition to existing data, we sought to confirm whether an apparent beneficial effect in cognitive function is observed after soy consumption in equol producers compared with non‐producers. Methods The present study was a cross‐sectional, observational study of 152 (male/female = 61/91, mean age 69.2 ± 9.2 years) individuals. Participants were divided into two groups according to equol production status, which was determined using urine samples collected after a soy challenge test. Cognitive function was assessed using two computer‐based questionnaires (touch panel‐type dementia assessment scale [TDAS] and mild cognitive impairment [MCI] screen). Results Overall, 60 (40%) of 152 participants were equol producers. Both TDAS and prevalence of MCI were significantly higher in the equol producer group than in the non‐producer group. In univariate analyses, TDAS significantly correlated with age, serum creatinine, estimated glomerular filtration rate and low‐density lipoprotein cholesterol. In multiple regression analysis using TDAS as a dependent variable, equol producer (β = 0.236, P = 0.005) was selected as an independent variable. In addition, multiple logistic regression analysis to assess the presence of MCI showed that being an equol producer was an independent risk factor for MCI (odds ratio 3.961). Conclusions Compared with equol non‐producers, equol producers showed an apparent beneficial effect in cognitive function after soy intake. Geriatr Gerontol Int 2017; ••: ••–••.