We aimed to explore the association between visit-to-visit systolic blood pressure variability (BPV) and cognitive function in individuals with type 2 diabetes. We performed a post hoc analysis of the Action to Control Cardiovascular Risk in Diabetes Memory in Diabetes (ACCORD-MIND) substudy. A total of 2867 diabetes patients with ≥3 BP measurements between the 4- and 20-month visits were included. Visit-to-visit systolic BPV was calculated. Cognitive decline was defined as a Mini-Mental State Exam (MMSE), Digit Symbol Substitution Test (DSST), or Rey Auditory Verbal Learning Test (RAVLT) score greater than 1 standard deviation (SD) below the baseline mean, or a Stroop test score more than 1 SD above the baseline mean. The associations of systolic BPV with risks of cognitive decline were examined using Cox proportional hazards models, and with changes in brain magnetic resonance imaging parameters were evaluated using mixed models. The risk of cognitive decline defined by the DSST score (but not by other scores) increased significantly with systolic BPV quartiles (p for trend = 0.008), and there was a 55% increased risk for BPV quartile 4 versus quartile 1 (hazard ratio = 1.55, 95% confidence interval 1.10-2.19). Furthermore, a positive correlation was observed between systolic BPV and change in white matter lesion volume (β = 0.07, 95% CI 0.01-0.13). A greater visit-to-visit systolic BPV was significantly associated with an increased risk of cognitive decline measured by DSST and an increase in white matter lesion volume in patients with type 2 diabetes.