Sumario:Background: Hyperhomocysteinemia is considered an independent risk factor for vascular occlusive diseases. To date, there is no general agreement on hyperhomocysteinemia cutoff values. Methods: To establish a homocysteine cutoff value, we performed a case-control study in 118 patients suffering from venous thrombosis and in 115 healthy subjects. We calculated odds ratios at different cutoff points and considered hyperhomocysteinemia as homocysteine levels above which the risk of venous thrombosis was increased. Results: Initially we used the 97.5th percentiles for fasting homocysteine levels in the control group to calculate odds ratios (95% CI) of 9.5 (2.6-35.3), 3.7 (0.8-17.9) and 4.5 (1.7-123.8) for the total population, women and men, respectively. When individuals with well-known thrombotic risk factors were excluded (selected population), odds ratios were 10.5 (2.7- 41.1), 6.5 (1.3-32.1) and 11.2 (1.2-103.1), respectively, confirming hyperhomocysteinemia as an independent risk factor for venous thrombosis. We did not find any association of venous thrombosis with the homozygous methylenetetrahydrofolate reductase C677T mutation. When the hyperhomocysteinemia cutoff was set at other arbitrary points, odds ratios for the selected population were statistically significant only at >12 μmol/L. Conclusions: Based on our results, we propose 12 μmol/L as the hyperhomocysteinemia cutoff value. © 2007 by Walter de Gruyter.