INTRODUCTION: Case-control studies suggest that elevated lipoprotein (a) (Lp(a)) is a risk factor for first venous thromboembolism (VTE). Lp(a) has not been prospectively investigated as a possible risk factor for recurrent VTE in first unprovoked VTE patients. We sought to determine if serum Lp(a) levels in patients with unprovoked VTE who discontinue anticoagulants after 5 to 7 months of therapy predict VTE recurrence in a prospective cohort study. MATERIALS AND METHODS: Serum Lp(a) measurements were obtained from 510 first unprovoked VTE patients treated for 5 -7 months with anticoagulants in a 12 center study. Patients were subsequently followed for a mean of 16.9 months (SD+/-11.2) for symptomatic VTE recurrence which was independently adjudicated with reference to baseline imaging. RESULTS: There was no significant association between Lp(a) as a continuous variable and recurrent VTE nor in gender stratified subgroups. No statistically significant differences were observed in the median Lp(a) concentrations between patients who recurred and those who did not recur (median (interquartile range): 0.09 g/L (0.17) versus 0.06 g/L (0.11) respectively; p=0.15). The Lp(a) cut-off point of 0.3g/L was not significantly associated with recurrent VTE for the overall population nor in gender stratified subgroups. CONCLUSIONS: Elevated serum Lp(a) does not appear to be associated with recurrent VTE in patients with history of first unprovoked VTE and may not play a role in identifying patients with unprovoked VTE at high risk of recurrence. There was no optimal predictive threshold for the overall population or for sex sub-groups and Lp(a) ≥ 0.3 g/L was not a significant predictor of recurrent VTE.