Obesity and physical activity (PA) may affect inflammation and are also related to depression. This study aimed to explore the association between depression, obesity, and physical activity (PA) in older diabetes patients mediated by inflammation. We conducted a cross-sectional study with 197 elderly diabetes patients (≥65 y/o). Participants were interviewed to gather demographic and lifestyle data. Assessment of depression was based on the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) criteria. High-sensitivity C-reactive protein was used as a marker of inflammation. Participants with a body mass index (kg/m2) ≥ 27 were considered to be obese. Our data indicated that among all participants with (n = 57) and without (n = 140) depression, older diabetes patients with depression had a lower intake of energy and protein and a lower prevalence of smoking and alcohol consumption than those without depression (p < 0.05). We also found that inflammation may be a partial mediator in the relationship between obesity and depression, and a significant mediator between PA and depression. Additionally, a regression model of obesity and PA showed that PA was a significant predictor of inflammation. However, the association between obesity and inflammation was not significant. When obesity, PA, and inflammation were included in a regression model together, inflammation significantly predicted depression (OR = 4.18, p = 0.004). The association between obesity and depression was also significant (OR = 2.45, p = 0.038). However, the association between PA and depression was not significant, and the mediating effect of inflammation was significant according to the Sobel test (z = −2.01, p = 0.045). In conclusion, the beneficial effects of PA may lower levels of inflammation produced by obesity, thus reducing inflammatory effects that may be related to depression. Overall, inflammation may mediate the relationship between depression and PA in older diabetes patients.