BACKGROUND Microalbuminuria is regarded as the most important predictor of high risk for the development of diabetic nephropathy. Early detection may allow treatment to prevent progression to persistent albuminuria and renal failure. Recent studies have shown that conventional immunoassays underestimate urinary albumin concentration, as albumin in urine may exist in two forms, immuno-reactive and immuno-unreactive. The present study examines the differential lead-time for the development of microalbuminuria as measured by both conventional radioimmunoassay (RIA; measures immuno-reactive) and high-performance liquid chromatography (HPLC; measures total albumin = immuno-reactive plus immuno-unreactive) analysis in both type 1 and type 2 diabetic patients. METHODS Analysis was performed on 511 stored urine samples collected over the last 13 years from type 1 diabetic patients who either progressed from normo- to microalbuminuria (progressors, N= 17), or who remained normoalbuminuric (nonprogressors, N= 25) as defined by RIA, and on 634 urine samples collected from patients with type 2 diabetes defined as either progressors (N= 24) or nonprogressors (N= 25). RESULTS For type 1 progressors, the mean lead-time for the HPLC assay versus the RIA was 3.9 years, with a 95% CI of 2.1 to 5.6 years. For type 2 progressors, the mean lead-time was 2.4 years with a 95% CI of 1.2 to 3.5 years. There was no significant difference between the lead-time analysis between type 1 and type 2 diabetic patients. CONCLUSIONS These results demonstrate that measurement of total albumin may allow earlier detection of microalbuminuria associated with diabetic nephropathy.