Human cystatin C is a low molecular mass protein of 13359 Dalton recently proposed as a new very sensitive marker of changes in glomerular filtration rate. Serum cystatin C concentration correlates negatively with glomerular filtration rate as well as or better than creatinine. We evaluated a recently introduced automated nephelometric immunoassay for cystatin C in serum or EDTA-plasma samples on the Behring Nephelometer System. The assay consists of incubating the 100-fold diluted sample for 6 minutes with latex particles covalently coated with anti-human cystatin C antibodies, and then quantifying the change of light-scatter produced. Method reproducibility is satisfactory, the intra- and inter-assay coefficients of variation ranging from 1.58% to 3.77% and from 5.6% to 11.47% respectively. Rheumatoid factor (< or = 1116 IU/ml), bilirubin (< or = 418 micrommol/l), triglycerides (10.47 mmol/), and haemoglobin (12 g/l) do not significantly interfere in the assay. No significant difference was found in cystatin C concentration between serum and EDTA-plasma samples. Cystatin C is stable in serum samples stored under different conditions up to one month. This method correlates well (mean difference=-0.536+/-0.307 mg/l) with another commercially available particle-enhanced turbidimetric immunoassay. Cystatin C offers better clinical sensitivity than creatinine for discriminating patients with normal renal function and those with mild-to-moderate reduction in renal function. This method is suitable for routine cystatin C measurement, including emergencies.