Background Blood pressure (BP) postural changes, both orthostatic hypotension (OHYPO) and orthostatic hypertension (OHYPER) are common in older adults. Few studies have investigated their association with cognition, particularly for OHYPER, an emerging cardiovascular risk factor. We aimed to assess the association between OHYPO, OHYPER and cognition in non‐institutionalized older subjects. Methods The S.AGES (Sujets ÂGES, Aged Subjects) cohort followed every 6 months for 3 years non‐institutionalized subjects aged ≥65 years without dementia at inclusion, in France. OHYPO and OHYPER were respectively defined as a fall or an increase of ≥20 mmHg in systolic BP and/or ≥10 mmHg in diastolic BP after standing from a sitting position. Cognition was assessed using the Mini‐Mental State Examination (MMSE). Linear mixed models were used for the analyses. Results Among the 3170 subjects included (mean age 78 years, 56% women), 209 (6.5%) had OHYPO and 226 (7.1%) had OHYPER at baseline. After adjustment for demographics, cardiovascular risk factors and disease, seated SBP/DBP and BP lowering treatment, mean MMSE was 0.52 point lower in participants with OHYPER compared to those with normal BP postural changes ( β adjusted [95% CI] = −0.52 [−0.96; −0.09], p = 0.02) and 0.50 point lower in participants with OHYPO compared to those with normal BP postural changes ( β adjusted [95% CI] = −0.50 [−0.95; −0.06], p = 0.03). Sensitivity analyses showed a dose–response relationship between OHYPO and cognition. Conclusion Although the absolute differences in MMSE were small, both OHYPO and OHYPER were associated with lower cognition. Orthostatic BP measurements could help identify patients with risk of cognitive impairment. Further studies are needed to assess whether controlling orthostatic BP could be a promising interventional target in preserving cognition among older adults.