P-selectin glycoprotein ligand-1 (PSGL-1) binding to Pselectin controls early leukocyte rolling during inflammation. Interestingly, antibodies and pharmacological inhibitors (eg, rPSGL-Ig) that target the N-terminus of PSGL-1 reduce but do not abolish P-selectin-dependent leukocyte rolling in vivo whereas PSGL-1-deficient mice have almost no P-selectin-dependent rolling. We have investigated mechanisms of P-selectin-dependent, PSGL-1-independent rolling using intravital microscopy. Initially we used fluorescent microspheres to study the potential of L-selectin and the minimal selectin ligand sialyl Lewis x (sLe x ) to interact with postcapillary venules in the absence of PSGL-1. Microspheres coated with combinations of L-selectin and sLe x interacted with surgically stimulated cremaster venules in a P-selectin-dependent manner. Microspheres coated with either L-selectin or sLe x alone showed less evidence of interaction. We also investigated leukocyte rolling in the presence of PSGL-1 antibody or inhibitor (rPSGL-Ig), both of which partially inhibited P-selectin-dependent leukocyte rolling. Residual rolling was substantially inhibited by L-selectin-blocking antibody or a previously described sLe x mimetic (CGP69669A). Together these data suggest that leukocytes can continue to roll in the absence of optimal P-selectin/PSGL-1 interaction using an alternative mechanism that involves P-selectin-, L-selectin-, and sLe and present attractive targets for anti-inflammatory drugs. P-Selectin may be a good target for cardiovascular disease therapy, 2-5 although notable clinical trial disappointments 6,7 suggest that effective P-selectin inhibition may be more challenging than expected.Natural ligand mimicry is a common drug development approach. Elements of PSGL-1 required for high-affinity P-selectin recognition include a sialylated, fucosylated O-glycan and tyrosine sulfation at appropriate positions near the N-terminus. 8 Drugs mimicking one or more of these elements could theoretically inhibit P-selectin. Inhibitors based on the carbohydrate selectin ligand, sLe x , are effective, albeit in high doses, against E-selectindependent leukocyte rolling in vivo but have no measurable effect on established P-selectin-dependent rolling.