Mild or moderate hyperhomocysteinemia as a risk factor for venous thrombosis is still a matter of debate. The strength of this study is to bring a body of elements to evaluate whether hyperhomocysteinemia should be used as a biomarker for venous thromboembolism (VTE). These elements consist of a biological evaluation of several hematological risk factors, and an original control group made of patients with a negative Doppler ultrasonography. A total of 151 cases and 155 controls were included. Total plasma homocysteine level, MTHFR C677T polymorphism, inherited abnormalities of the natural anticoagulant system as well as plasma folate and cobalamin levels were determined. A total of 41 (27.2 %) of cases and only 9 (5.8%) of controls had at least one of the coagulation defects studied. No significant difference was observed for total homocysteine levels between the 2 groups: median (interquartile range) = 8.3 (7.2-10.8) micromol/L for cases and 8.4 (7-10.9) micromol/L for controls. We found significantly more plasma folates and/or cobalamin deficiencies in controls (18.3%) than in cases (8.6%). After adjustment for several variables significantly related to risk factors of VTE, hyperhomocysteinemia (>13.2 micromol/L) was not found statistically associated with VTE: odds ratio 1.36 (95% confidence interval, 0.52-3.54). The prevalence of the homozygous 677TT polymorphism in the MTHFR gene was not increased in cases compared with controls. Mild or moderate hyperhomocysteinemia does not seem to be a strong determinant in VTE not only when the control group does not exclusively include healthy persons but also in investigated disease-free (thromboembolic disease) controls.