Heparin and low molecular weight heparin (LMWH) are widely used for treatment of cancer patients with thrombosis, a common complication of malignant disease. Several recent prospective clinical studies indicate that heparin might improve outcomes of human cancer. Meanwhile, experimental evidence from mouse models consistently demonstrates that heparin efficiently inhibits metastasis. We have previously shown that P- and L-selectin play independent roles in supporting the initial stages of hematogeneous metastasis. Heparin is a known potent inhibitor of such selectin-mediated interactions. Here we provide evidence that the absence of both P- and L-selectin (PL -/- mice) dramatically improved survival in an experimental metastasis model. The use of clinically acceptable amounts of heparin did not further aff5ct metastasis rates in such mice. However, a non-anticoagulant derivative of heparin with P- and L-selectin inhibitory properties reduced metastasis to similar levels as observed in PL -/- mice. The virtual elimination of metastasis by a single treatment with a modified heparin without anticoagulant activity strongly suggests that heparin primarily reduces metastatic disease by inhibiting P- and L-selectin interactions. However, such heparins could have further effects at higher doses.