The aim of this randomized controlled trial was to determine the effects of 8-week exercise-intervention on cognition and related serum biochemical markers in nonagenarians. We also studied the effects of a 4-week training cessation (‘detraining’) period on our study variables. Participants were randomly allocated to a standardcare (control) or intervention (exercise) group [n = 20 (16 women)/group]. The intervention focused on supervised, light-to-moderate-intensity aerobic and resistance exercises (mainly leg press), and included 3 weekly sessions. Cognitive status was determined by the mini-mental state examination and geriatric depression scale. We analysed proteins with reported relation with mechanisms behind cognition changes such as serum levels of angiotensin converting enzyme, amyloid-precursor protein, epidermal growth factor, brain-derived neural factor and tumor necrosis factor. No significant change (P > 0.05) in any of the variables studied was found following the exercise intervention compared with the standard-care group. Similarly, no significant changes (P > 0.05) were observed following the detraining period compared with the standardcare group. Overall changes after the exercise intervention in serum biomarkers were not associated with changes in functional capacity and cognitive measures. An 8-week exercise intervention focusing on resistance exercises neither benefits cognitive function nor affects the levels of the serum proteins analysed in nonagenarians.