Use of whole blood in adhesion assays allows analysis of the rheological and haematological factors that may influence adhesion, and avoids the need for isolation procedures that may modify the properties of leucocytes. We have adapted an in vitro flow model to allow videomicroscopy of leucocytes fluorescently labelled with rhodamine 6G (20 μg/ml) in anticoagulated whole blood. Blood was perfused at a range of wall shear rates (35–280/s) through a vertical glass capillary with a rectangular cross-section (microslide) that had been coated with P-selectin (10 μg/ml). Nearly all adherent cells were rolling in blood that had been anticoagulated with buffered citrate, but 40–50% became immobilized when heparin or thrombin inhibitor (PPACK) were used. The efficiency of leucocyte adhesion decreased steadily during 1–4 h of blood storage. Smaller fluorescent cells (lymphocytes) adhered less efficiently than larger cells (granulocytes) and rolled faster. Adhesion decreased monotonically with increasing wall shear rate or stress, but the velocity of rolling varied little. Among healthy volunteer donors, adhesion correlated with blood leucocyte count, but did not vary significantly with natural variation in haematocrit, blood viscosity or red cell aggregation. In conclusion, we have characterized adhesion of leucocytes in flowing whole blood, identified key experimental variables and demonstrated that physical environmental factors can markedly influence adhesive behaviour.