Two automated turbidimetric D-dimer assays (BC D-dimer Plus, Dade Behring, Marburg, Germany and Auto-Dimer, Biopool, Umeå, Sweden) were compared to two enzyme-linked immunosorbent assays (ELISAs) (Enzygnost D-dimer micro, Dade Behring and Asserachrome D-dimer, Diagnostica Stago, Asnières, France) and two rapid D-dimer assays (SimpliRed, Agen Biomedical, Brisbane, Australia and Minutex, Biopool) in out-patients with suspected deep vein thrombosis (DVT). In addition, the performance of prothrombin fragment 1+2 (F1+2), thrombin-antithrombin complex (TAT), and soluble adhesion molecules VCAM-1 and P-selectin for DVT diagnosis was assessed. One hundred and thirty-five consecutive out-patients with suspected DVT of the lower limb were included, and in 52 (39%), DVT was confirmed by compression ultrasound. All D-dimer assays investigated reliably excluded DVT in those patients without DVT irrespective of their pre-test clinical probability of DVT. One D-dimer ELISA (Dade Behring) gave the highest area under the receiver operating characteristic (ROC) curve compared to other assays, and therefore, this was the most accurate assay in differentiating patients with from patients without DVT. The diagnostic performance of one automated turbidimetric assay (Auto Dimer, Biopool) was similar to ELISA and its convenience close to rapid latex agglutination assays. Most patients with a high pre-test clinical probability of DVT had positive D-dimer regardless of the presence or absence of DVT, which decreased the specificity of the tests and made D-dimer determination less useful for this group of patients. Because the diagnostic accuracy [sensitivity, specificity, negative (NPV) and positive predictive value (PPV)] of F1+2, TAT, VCAM-1 and P-selectin was inferior to D-dimer assay, these assays could not substitute or supplement D-dimer testing in diagnosis of DVT. Levels of VCAM-1 and P-selectin were increased in patients with DVT and should therefore be investigated further to clarify their role in DVT.