Objective: Evaluate whether pre-frail and frail PLWH have a higher risk of cognitive impairment on screens. Methods: Analysis of PLWH aged 70 or older included in the ANRS EP66 SEPTAVIH cohort, on antiretroviral therapy for at least 12 months and with a MoCA test at enrolment. Adjusted risk of a Montreal Cognitive Assessment (MoCA) 30.2), and 41.3% were college graduates; 27.3% had a history of clinical AIDS. 294 (58.5%) PLWH had a MoCA score <26; 182 (36%) a MoCA score ≤ 23. Frailty, pre-frailty, and robustness were found in 13.1%, 63.6% and 23.3% participants, respectively. PLWH with a MoCA < 26 had a significantly higher risk of being frail/prefrail, this before (OR = 2.31 [95% CI: 1.50, 3.57]), and after adjustment for confounders (OR = 1.80, [1.07–3.01]). The risk of being frail/prefrail in patients with a MoCA ≤ 23 was higher (adjusted OR = 2.75, [95%CI: 1.46–5.16]). Other factors independently associated with a MoCA < 26 were older age, birth outside of France, and a lower education level and being diabetic. Conclusions: Abnormal MoCA screens were frequent in our cohort of PLWH aged 70 or older with controlled HIV disease. Cognitive impairment should be systematically screened in frail/prefrail PLWH. Frailty/pre-frailty, diabetes and social factors, but not HIV-related factors, are important determinants of cognitive function in PLWH with controlled disease.Read less <