We prospectively evaluated the diagnostic performance of a new soluble fibrin assay in 303 consecutive patients with suspected pulmonary embolism and examined potentially useful cut-off levels at which this disease can be safely excluded. In addition, the diagnostic accuracy was calculated in the subgroups of in- and outpatients. The ROC curve of the assay in the total study cohort had an area under the curve of 0.69. The cut-off level associated with a sensitivity and negative predictive value of 100% was 20 ng/ml, but the specificity was only 4%. The cut-off level with a sensitivity of 90% was 30 ng/ml, which corresponded with a specificity and negative predictive value of 27% and 86%, respectively. The diagnostic performance was comparable in the subgroups of in- and outpatients. We conclude that the soluble fibrin assay has a low diagnostic accuracy and seems unsuitable as a screening test for the exclusion of pulmunary embolism.