Objective: The objective of our study was to quantify the extent to which age, sex, and education influence the cross-sectional results of four neuropsychological tests in a population aged 40 to 75 years. Methods: Cognitive functions were assessed in 25,748 volunteers aged 45 to 75 years, enrolled in the Constances cohort between January 2012 and May 2014. The scores from the neuropsychological tests used were (1) the free, immediate, and delayed recall scores from the "16-item Free and Cued Recall Test" (RLRI16), (2) semantic and phonemic verbal fluency tasks, (3) the "Digit Symbol Substitution Test" (DSST), and (4) parts A and B of the "Trail Making Test" (TMT-A and B). The strength of the associations was quantified using R² coefficients representing the percentages of score variation that could be explained by age, sex, and educational level. Results: age explained 0.6% of the variation in verbal fluency scores; this proportion rose to 8.1% for the TMT-A test. Gender accounted for 6.3% and 5.0% of the variation in the two RL/RI16 scores, 2.5% of that in the DSST, and 0.7% of that in phonemic fluency. Regarding education, educational level explained 1.8% of the variation in the TMT-A, and this proportion rose to 11.5% for the DSST. Ultimately, age, gender, and education, when considered together, explained between 8% (for semantic fluency) and 24% (for the DSST) of the variations in cognitive scores. Conclusion: age, gender, and education are strongly associated with cognitive performance.