The onset of diabetic nephropathy is characterised by a rise in albumin excretion rate (AER) and/or a transient rise in glomerular filtration rate (GFR) (hyperfiltration). Without intervention AER increases exponentially and there is a linear decrease in GFR after onset of overt nephropathy. In overt nephropathy, AER is a predictor of decline in GFR and the early AER response to antihypertensive therapy correlates with long-term decline in GFR. AER can be measured by immunoassay or by other techniques including HPLC. However, HPLC assays result in higher levels of AER in normal subjects compared with immunoassayable AER. Recent data suggest that there are distinct albuminuric and non-albuminuric pathways to renal impairment in type 1 and type 2 diabetes. In type 2 diabetes, the non-albuminuric pathway may explain a decline in GFR to <60 ml/min/1.73 m(2) in approximately one in four subjects after accounting for the use of renin angiotensin system inhibitors. In established nephropathy (chronic kidney disease (CKD) stages 3 and 4), plasma cystatin C based estimates of GFR are marginally superior to creatinine based estimates. However, cystatin C clearly outperforms creatinine based estimates of GFR decline at GFR levels >60 ml/min/1.73 m(2) (CKD stages 1 and 2). Other potential markers of progression of diabetic nephropathy include transforming growth factor beta (TGFbeta) and connective tissue growth factor (CTGF). However, long-term studies are needed to define their roles as markers of progression. Diabetic nephropathy is likely to be more susceptible to intervention at an early stage and accurate estimation of GFR is already possible with cystatin C. However, improved formulas for estimating GFR based on using creatinine or other markers are still required, because this may still provide the most cost effective approach applicable to existing clinical practice.