BACKGROUND: No longitudinal study has investigated the impact of combination of kidney function (KF) and physical function (PF) on cognitive decline in these patients. METHODS: We conducted a 2-year prospective cohort study enrolling 131 patients ≥ 65 years with pre-dialysis chronic kidney disease (CKD). We assessed cognitive function with the Japanese version of the Montreal Cognitive Assessment (MoCA-J). We calculated %MoCA-J based on the rate of change between baseline and follow-up MoCA-J scores, and defined cognitive decline over 2 years as a %MoCA-J of less than the first quartile value. We defined eGFR ≥ 30 as mild-to-moderate and eGFR < 30 mL/min per 1.73 m2 as severe. In addition, low PF was defined as low handgrip strength (< 26 for men and < 18 kgf for women) and/or low gait speed (< 0.8 m/s). Patients were classified into four groups: group 1, patients with mild-to-moderate impairment in KF and high PF; group 2, with mild-to-moderate impairment in KF and low PF; group 3, with severe impairment in KF and high PF; and group 4, with severe impairment in KF and low PF. RESULTS: Eighty-four patients completed follow-up assessment. Multivariate logistic regression analysis showed that the combination of severe impairment in KF and low PF was significantly associated with cognitive decline (odds ratio 5.73). However, no significant cognitive decline was observed in patients with either severe impairment in KF or low PF alone. CONCLUSIONS: We may need to focus on maintaining PF in older patients with advanced CKD may help to prevent cognitive decline.