Background: In this study, we evaluated the effects of 6 weeks of voluntary exercise on a running wheel and allium paradoxum on tau protein in the cerebellum of diabetic rats. Methods: After obtaining 42 adult male Wistar rats (8 weeks old), they were divided into control, diabetic, training, training-diabetic, training-diabetic-antioxidant, and diabetic-antioxidant groups. To induce diabetes, 120 mg/kg of alloxan (dissolved in saline) was intraperitoneally administered. Tau protein levels in the cerebellum tissue were evaluated using an enzyme-linked immunosorbent assay (ELISA) kit. Data was analyzed using one-way analysis of variance (ANOVA) and least significant different (LSD) post-hoc test. Findings: Induction of diabetes caused no significant changes in tau protein levels in the cerebellum. Allium paradoxum significantly increased tau protein levels in diabetic rats (P = 0.025). Moreover, receiving allium paradoxum resulted in significantly higher amounts of tau protein compared to voluntary exercise (P = 0.001) and the combination of antioxidant and exercise (P = 0.001). However, the training-diabetic and training-diabetic-antioxidant groups were not significantly different (P = 0.945). Conclusion: The results of this research showed that alloxan-induced diabetes does not alter the tau protein levels in cerebellum. In response to voluntary exercise on running wheel, tau protein content of the cerebellum did not significantly change. Allium paradoxum increased phosphorylated tau in cerebellum of diabetic rats. However, more studies are required to determine beneficial effects of physical exercise on tau phosphorylation. Keywords: Diabetes, Cerebellar tau protein, Allium paradoxum, Voluntary training, Rats