The surface expression of adhesion molecules and other function-associated antigens on peripheral blood leucocytes may be measured by flow cytometry. However, quantification of these antigens is difficult, because their expression can be rapidly and artefactually modulated if the cells are activated in vitro. Consequently, it is common, when analyzing these antigens either: (1) to label leucocytes in whole blood at 4 degrees C, to lyse erythrocytes and then fix the leucocytes, or (2) to fix the leucocytes in whole blood, to lyse the erythrocytes, and then label the leucocytes. We have compared the mean fluorescence intensity (MFI) values for CD11b, CD18, and L-selectin (Leu-8 and TQ1 epitopes) on human peripheral blood leucocytes, using these two approaches. In addition, we have simultaneously evaluated how anticoagulants (acid citrate, K3EDTA, and heparin) and the presence or absence of divalent metal ions (Ca2+ and Mg2+) affect the expression levels of these antigens. The results for all four epitopes varied markedly depending on the preparation procedure used but were less affected by the choice of anticoagulant and whether divalent cations were or were not present in the media used for cell preparation and labelling. Comparison of the results obtained using these procedures, which involve fixation with formaldehyde, with those obtained by a recently developed procedure in which unfixed leucocytes were labelled with the vital nuclear dye LDS-751 and antibodies together, then analysed in unlysed whole blood at 4 degrees C, showed that formaldehyde-based preparation techniques underestimated the expression (MFI) of CD18, Leu-8, and TQ1. It is recommended that, whenever practicable, measurements are made on unfixed cells stained using the newer procedure.