Heparin treatment of venous thromboembolic disease has been validated since 1960. Nevertheless no study was sufficient to determine an optimal therapeutic schedule between sub-cutaneous (SC) unfractionated heparin (UFH), intravenous (IV) UFH and low molecular weight heparin (LMWH). One meta-analysis showed a significant risk reduction of recurrent thromboembolic events (OR = 0.58, CI 95 per cent [0.34-0.99]) and a non-significant risk reduction of haemorrhagic events (OR = 0.78 [0.40-1.52]) with UFH SC compared to UFH IV, but homogeneity testing was significant (p < 0.001). Some discrepancy was shown between the results of the three metaanalyses which compared LMWH to UFH according to the selection criteria of clinical trials used. With an exhaustive selection, LMWH involved a non-significant risk reduction of recurrent thromboembolic events (OR = 0.66 [0.41-1.07], p = 0.09), and a non-significant risk reduction of haemorrhagic events (OR = 0.65 [0.36-1.16], p = 0.15). So no definitive conclusion could be drawn but it seems that UFH can be recommended whatever the administration route or LMWH for deep vein thrombosis treatment.