Microalbuminuria is still the only early abnormality of the diabetic kidney that has an established prognostic value. Microalbuminuria evolves into clinical nephropathy and renal failure in a majority of cases of insulin-dependent diabetic patients, and is defined by the detection of urinary albumin excretion rates of 20-200 microg/min in timed urine collections. The occurrence of microalbuminuria at rates of 5-27 % of non-proteinuric patients and cost-benefit considerations justify the screening for microalbuminuria in diabetic outpatient clinics. Both near-normalisation of glycaemic control and treatment with ACE-inhibitors are indicated in patients with insulin-dependent diabetes to correct the progression of micro- to macroalbuminuria. Other therapeutic perspectives are being considered, but the current notion that the available therapies may not arrest the course of nephropathy at this stage suggests that earlier interventions may be required. Prevention of microalbuminuria and overt nephropathy may require a primary approach to the subset of patients with a genetic predisposition to this complication, and several studies (candidate gene or genomic scan with microsatellite probes) now address the chromosomal loci and the nature of the genes that may be involved.