Clinical assessment of glomerular filtration rate (GFR) mainly relies on single determinations of serum creatinine (crea) which is commonly insensitive to mild renal dysfunction. Serum cystatin C (cysC) has been proposed as an alternative endogenous marker of GFR showing higher correlation to standard clearance methods such as inulin or iohexol clearance. We compared serum crea and cysC levels in n=127 patients undergoing cardiac catheterization. The clearance of the iodinated contrast dye iopromide served as reference method for GFR. Serum cysC was determined by a particle-enhanced immunonephelometric method. CysC showed higher non-parametric correlation (r=0.805) to the iopromide clearance compared to crea (r=0.652) and to the estimated GFR according to the Cockcroft-Gault formula (r=0.690), which underestimated true GFR systematically. Receiver operating curves revealed a greater area-under-the-curve (AUC) for cysC (0.957 vs. 0.801, p<0.05). At a cut-off level of >1.3 mg/l cysC exhibited an 88% sensitivity and a 96% specificity for detecting renal dysfunction which was defined as an iopromide clearance less than 80 ml/min/1.73 m2; best values for crea were 63% for sensitivity and 80% for specificity at a cut-off of >1.2 mg/dl. In conclusion, cysC detected reduced GFR more reliably and at an earlier stage in patients undergoing cardiac catheterization allowing a better identification of patients with renal dysfunction and those at risk for contrast damage.