Albuminuria is an early marker for diabetic nephropathy in patients with diabetes, and has a clear place in patient care. It also predicts cardiovascular events and mortality in diabetic patients and in the general population, and is slowly becoming accepted in population screening for cardiovascular disease and chronic kidney disease. Recently, investigators found that a considerable amount of albumin in urine is nonimmunoreactive and that classic immunochemical assays do not properly measure all albumin in urine. Assays that detect immunoreactive plus nonimmunoreactive albumin may better predict development of diabetic nephropathy, cardiovascular events, and mortality than assays that only detect immunoreactive albumin. Proof of the existence of nonimmunoreactive albumin emerged from the finding that albumin contains urine fragments. In this review, we critically appraise the presence and relevance of albumin fragments and nonimmunoreactive albumin molecules in urine, and the potential additive value of albuminuria detected by assays that assess nonimmunoreactive plus immunoreactive albumin over albuminuria detected by classic immunochemical assays in predicting end points.