BACKGROUND Extremely low frequency pulsed magnetic fields (ELFPMF) have been shown to induce Faraday currents and measurable effects on biological systems. A kind of very high frequency electromagnetic field was reported that it improved the symptoms of diabetic nephropathy (DN) which is a major complication of diabetes. However, few studies have examined the effects of ELFPMF DN at the present. The present study was designed to investigate the effects of ELFPMF on DN in streptozotocin (STZ)-induced type 1 diabetic rats. METHODS Adult male SD rats were randomly divided into three weight-matched groups: Control (non-diabetic rats without DN), DN + ELFPMF (diabetic rats with DN exposed to ELFPMF, 8 h/days, 6 weeks) and DN (diabetic rats with DN exposed to sham ELFPMF). Renal morphology was examined by light and electron microscopy, vascular endothelial growth factor (VEGF)-A and connective tissue growth factor (CTGF) were measured by enzyme linked immune sorbent assay. RESULTS After 6 weeks' ELFPMF exposure, alterations of hyperglycemia and weight loss in STZ-treated rats with DN were not found, while both positive and negative effects of ELFPMF on the development of DN in diabetic rats were observed. The positive one was that ELFPMF exposure attenuated the pathological alterations in renal structure observed in STZ-treated rats with DN, which were demonstrated by slighter glomerular and tubule-interstitial lesions examined by light microscopy and slighter damage to glomerular basement membrane and podocyte foot processes examined by electron microscopy. And then, the negative one was that ELFPMF stimulation statistically significantly decreased renal expression of VEGF-A and statistically significantly increased renal expression of CTGF in diabetic rats with DN, which might partially aggravate the symptoms of DN. CONCLUSIONS Both positive and negative effects of ELFPMF on the development of DN in diabetic rats were observed. The positive effect induced by ELFPMF might play a dominant role in the procession of DN in diabetic rats, and it is suggested that the positive effect should be derived from the correction of pathogenic diabetes-induced mediators.