Slight albuminuria predicts clinically significant nephropathy in patients with insulin-dependent diabetes mellitus (IDDM) and early death in patients with non-insulin-dependent diabetes mellitus (NIDDM). This study compares four commercially available methods for measuring low concentrations of urinary albumin. We tested random spot urine specimens from 50 nondiabetic volunteers and 100 diabetic patients. This specimen was chosen to simplify collection in an outpatient setting. Two screening methods were evaluated for their ability to detect urinary albumin in the range of 15 to 200 mg/L. Sensitivity and specificity were 100% and 54.7%, respectively, for the Ames Micro-Bumintest; and 95.5% and 91.5%, respectively, for the Sclavo Albumin Screen. The high number of false-positive results made the Micro-Bumintest unacceptable. The Albumin Screen yielded fewer false-positive results, but also produced some false-negatives. Two quantitative methods, a radioimmunoassay (RIA) and a turbidimetric assay (the SPQ Microalbumin), yielded results that agreed well with each other.