BACKGROUND: Decline in cognitive performance is a highly prevalent health condition in elderly. We studied whether offspring of nonagenarian siblings with a familial history of longevity, perform better on cognitive tests compared to their partners as controls. This is relevant since it could provide insights into determinants underlying decline in cognitive performance. METHODS: Cross-sectional analysis within the longitudinal cohort of the Leiden Longevity Study consisting of middle-aged offspring of nonagenarian siblings together with their partners (n = 500, mean age (SD) 66.3 (6.1) and 65.7 (7.2) years, respectively) as controls. Memory function, attention and processing speed were tested using the 15-Picture Learning Test, Stroop test and Digit Symbol Substitution Test. Data were analyzed with regression adjusted for age, gender, years of education and additionally for diabetes mellitus, cardiovascular diseases, alcohol use, smoking, inflammatory markers and apolipoprotein E genotype. Robust standard errors were used to account for familial relationships among the offspring. RESULTS: Cognitive performance was worse at higher calendar age (p<0.001, all except Stroop test part 1). The offspring performed better compared to their partners on trial 3 (p = 0.005), the immediate (p = 0.016) and delayed (p = 0.004) recall of the 15-Picture Learning Test as well as on the interference and combined interference score of the Stroop test (p = 0.014 and p = 0.036, respectively) in the fully adjusted model. The difference between offspring and partners was estimated to be more than three years according to the observed difference in calendar age. CONCLUSIONS: Offspring of nonagenarian siblings with a familial history of longevity have better cognitive performance compared to the group of their partners of comparable age. This effect is independent of age-related diseases and known possible confounders. Possible explanations might be differences in subclinical vascular pathology between both groups.