<b><i>Objective:</i></b> This study was designed to evaluate whether chronic <i>Rosa canina</i> (RC) extract administration could improve recognition memory and depressive-like behavior in diabetic mice. <b><i>Materials and Methods:</i></b> Seventy-five male albino mice (25-30 g) were randomly divided into 5 groups (15 in each group). A single intraperitoneal injection of 200 mg/kg streptozotocin (STZ) was administered to the mice to induce diabetes. The control group received normal saline, and the diabetic groups received normal saline or 50, 250, and 500 mg/kg of RC extract for 28 days. The mice were weighed each week. Recognition memory and depressive-like behavior were assessed using forced swimming and novel object recognition (NOR) tests, respectively. Malondialdehyde (MDA) levels and total antioxidant capacity (TAC) were measured in the mouse brain homogenate to evaluate oxidative stress. Statistical analysis was conducted using SPSS, version 22. <b><i>Results:</i></b> The groups receiving 250 or 500 mg/kg RC had significantly lower immobility time (159.4 ± 4.7 and 150.1 ± 3.1 s) compared to the sham control group (192.1 ± 7.8 s) in the forced swimming test, and a higher discrimination index (0.39 ± 0.02 and 0.48 ± 0.03) was seen in diabetic animals in the NOR task compared to the sham control group (0.2 ± 0.01). Also, the groups receiving treatment with RC (250 and 500 mg/kg) had significantly higher TAC (0.92 ± 0.04 and 0.96 ± 0.05 mmol/L) and lower MDA (0.76 ± 0.02 and 0.67 ± 0.03 nmol/mg protein) levels in the brains in comparison to the model group. In the 3rd and 4th weeks of study, the RC-treated mice (250 and 500 mg/kg) gained more weight (31.2 ± 0.3 and 32.4 ± 0.3 g, and 31.3 ± 0.2 and 33.7 ± 0.3 g, respectively) than the diabetic group (30 ± 0.2 and 29.6 ± 0.3 g). <b><i>Conclusion:</i></b> This study showed that RC attenuated impairment of recognition memory and depressive-like behavior probably through modulation of oxidative stress in an STZ model of diabetes in mouse brains.