Evidence suggests that dietary walnuts are able to induce improvements in memory and learning functions. In addition, polyphenols have been shown to modulate critical neuronal signalling pathways involved in processes of learning and memory. The aim of our present work was to study the effect of polyphenol extracts from walnut testa (42%) on learning and memory functions in hypercholesterolemia mice based on obesity, hypercholesterolemia and oxidative stress. At the beginning of the experiment, mice were divided into 3 groups, one of them served as normol control group (NCG), the second as hypercholesterolemia control group (HCG), the last as walnut polyphenol-treated group (WTG). After 8 weeks of treatment, we investigated the performance of C57BL/6J mice in Morris water maze test. The results showed that the escape latency was significant increase in HCG, when compared to NCG and WTG. In addition, the number of crossings was significant decrease in HCG, when compared to NCG and WTG (-47.35% and -43.35%, respectively, P < 0.01). The swimming distance was longer in HCG than NCG and WTG [F (2,96) = 44.45, P < 0.001]. However, there was no significant difference in swimming speed among NCG, HCG and WTG [F (2, 96) = 0.167, P > 0.05]. On the other hand, walnut polyphenol (WP) significantly decreased serum total triglycerides, cholesterol and malondialdehyde (MDA) level (-36.31%, -31.48% and -21.51%, respectively, P < 0.01) and increased superoxide dismutase (SOD) activity (+48.39%, P < 0.01). Administration of WP significantly decreased MDA level (-36.86%, P < 0.01) and increased SOD activity (+17.08%, P < 0.01) in brain tissues. In conclusion, walnut polyphenol was able to improve learning and memory functions.