This study aimed to determine the effect of long-term exercise on the risk of developing cardiovascular diseases (CVD) and all-cause mortality in patients with diabetic kidney disease (DKD). A single-center, prospective intervention study using propensity score matching was performed over 24 months. The intervention group (n = 67) received six months of individual exercise instruction from a physical therapist, who performed aerobic and muscle-strengthening exercises under unsupervised conditions. New events were defined as the composite endpoint of stroke or CVD requiring hospitalization, initiation of hemodialysis or peritoneal dialysis, or all-cause mortality. The cumulative survival rate without new events at 24 months was significantly higher in the intervention group (0.881, p = 0.016) than in the control group (n = 67, 0.715). Two-way analysis of variance revealed a significant effect of the group factor on high density lipoprotein-cholesterol (HDL-C) which was higher in the intervention group than in the control group (p = 0.004); eGFRcr showed a significant effect of the time factor, which was lower at 24 months than before intervention (p = 0.043). No interactions were observed for all items. In conclusion, aerobic exercises combined with upper and lower limb muscle strengthening for six months reduce the risk of developing CVD and all-cause mortality in patients with DKD.