P-selectin glycoprotein-1 (PSGL-1) supports P-selectin-dependent rolling in vivo and in vitro. However, controversy exists regarding the importance of PSGL-1-dependent and -independent E-selectin rolling. Using antibodies against PSGL-1 and PSGL-1(-/-) mice, we demonstrated abolition of P-selectin-dependent rolling but only partial inhibition of E-selectin-mediated rolling in the cremaster microcirculation following local administration of tumor necrosis factor alpha (TNF-alpha). In vitro studies demonstrated that binding of recombinant mouse E-selectin chimera to PSGL-1(-/-) neutrophils was dramatically decreased in mice treated systemically but not locally with TNF-alpha. Further, PSGL-1 blockade abolished E-selectin-dependent rolling in wild-type mice following systemic TNF-alpha administration but not local TNF-alpha administration. Together, these data support an E-selectin ligand present on PSGL-1(-/-) neutrophils that is down-regulatable upon systemic but not local activation. To determine whether the PSGL-1-independent E-selectin ligand was physiologically important, we used a P- and E-selectin-dependent cutaneous contact hypersensitivity model. Binding studies showed no E-selectin ligand down-regulation in this model. The few cells that rolled on E-selectin ligand following PSGL-1 antibody administration or in PSGL-1 deficiency were sufficient to induce profound contact hypersensitivity. In conclusion, E-selectin mediates PSGL-1-dependent and independent rolling and the latter can be down-regulated by systemic activation and can replace PSGL-1 to support the development of inflammation.