Abstract Introduction As 30 to 50% of deep venous thrombosis (DVT) cases remain idiopathic, an increased focus on hematologic variables may therefore reveal novel correlates of DVT. Very few studies have investigated the association of hematological parameters with DVT and the causal relationship between them is still to be elucidated. Therefore, we aimed to investigate the association between serial values of hematologic variables and DVT. Methods Complete blood count parameters were serially measured at baseline and then at approximately 3-month intervals for 12 months in 152 adults with the first episode of DVT and 152 age- and sex-matched controls. The odds ratio (OR) with the 95% confidence interval (95%CI) was calculated as a measure of association between hematological parameters and DVT. Results The red cell distribution width (RDW) was the only hematologic variable which showed an independent and consistent association with DVT at all time points (multivariable-adjusted OR [95%CI] 3.38 [1.28 - 8.91] at baseline, 2.24 [0.85 - 5.92] at 3 months and 2.12 [0.81 - 5.55] at 12 months for RDW > 14.0%). This association was higher for provoked DVT than unprovoked DVT and for DVT plus pulmonary embolism than DVT alone. No significant correlation was found between the high RDW and classical thrombotic risk factors, except malignancy. Conclusions We demonstrated an independent and consistent association of the high RDW with the first episode of DVT in adult patients. The study was probably underpowered to evaluate the association between the high RDW and recurrent DVT. Further large studies with long follow-up are needed to confirm this association.