Abstract Background While midlife hypertension is deleterious, late-life hypertension has been associated with better cognitive outcomes in several studies. Many questions remain, including the relative benefit or harm of a blood pressure (BP) target and antihypertensive therapy of <120 in very old individuals. Methods The Sydney Memory and Aging Study (n = 1015) comprises a cohort of 70- to 90-year-olds, who were followed biennially for 8 years. Global cognition was assessed with a battery of 10 neuropsychological tests. Blood pressure was measured at Waves 1 and 2 and classified into 3 systolic groupings: group 1 (≤120 mmHg), group 2 (121–140 mmHg), and group 3 (>140 mmHg). Multiple regression, linear mixed modeling, and Cox regression examined the effect of BP and antihypertensives. Results There were no overall significant differences in global cognition or dementia between the disparate BP groups. However, in those not taking antihypertensives, the systolic BP (SBP) > 140 mmHg group had a significantly worse global cognitive trajectory compared to SBP ≤ 120 mmHg (b = −0.067, 95% CI [−0.129, −0.006], p = .030). Within the SBP ≤ 120 mmHg group those taking antihypertensives had significantly worse global cognition trajectories compared to those not taking antihypertensives even when controlling for past history of hypertension (b = −0.077, 95% CI [−0.147, −0.007], p = .030). Conclusions Untreated hypertension in old age is related to worse global cognitive decline. However, ongoing treatment at new recommendations of lower SBP targets may be related to poorer cognitive decline and should be considered carefully in older populations.