Diabetic nephropathy is the most common cause of end-stage renal disease. Therefore, novel therapies for the suppression of diabetic nephropathy must be developed. Rodent models are useful for elucidating the pathogenesis of diseases and testing novel therapies, and many type 1 and type 2 diabetic rodent models have been established for the study of diabetes and diabetic complications. Streptozotocin (STZ)-induced diabetic animals are widely used as a model of type 1 diabetes. Akita diabetic mice that have an Ins2+/C96Y mutation and OVE26 mice that overexpress calmodulin in pancreatic β-cells serve as a genetic model of type 1 diabetes. In addition, db/db mice, KK-Ay mice, Zucker diabetic fatty rats, Wistar fatty rats, Otsuka Long-Evans Tokushima Fatty rats and Goto-Kakizaki rats serve as rodent models of type 2 diabetes. An animal model of diabetic nephropathy should exhibit progressive albuminuria and a decrease in renal function, as well as the characteristic histological changes in the glomeruli and the tubulointerstitial lesions that are observed in cases of human diabetic nephropathy. A rodent model that strongly exhibits all these features of human diabetic nephropathy has not yet been developed. However, the currently available rodent models of diabetes can be useful in the study of diabetic nephropathy by increasing our understanding of the features of each diabetic rodent model. Furthermore, the genetic background and strain of each mouse model result in differences in susceptibility to diabetic nephropathy with albuminuria and the development of glomerular and tubulointerstitial lesions. Therefore, the validation of an animal model reproducing human diabetic nephropathy will significantly facilitate our understanding of the underlying genetic mechanisms that contribute to the development of diabetic nephropathy. In this review, we focus on rodent models of diabetes and discuss the utility and limitations of these models for the study of diabetic nephropathy.