The flow cytometric analysis of leucocytes in whole blood is usually performed on samples in which the erythrocytes have been lysed and the leucocytes fixed. Because lysis and fixation reagents have the potential to introduce artefacts, several commercially available reagents were used to prepare normal and leukaemic lymphocytes for immunophenotypic analysis by flow cytometry, and the results were compared with those obtained from live whole blood. The reagents tested were the ImmunoPrep system and OptiLyse C (Coulter), LF‐1000‐Lyse and Flow (Harlan), Uti‐Lyse (Dako) and FACS Lysing Solution (Becton Dickinson). The effect of each reagent on the apparent expression of CD3, CD5, CD11b, CD45, FMC7, κ and λ antigens was determined on lymphocytes from six normal controls and from six patients with chronic lymphocytic leukaemia (CLL). The following observations were made: (i) the time in minutes for each procedure varied markedly and was 1.5, 15, 20, 30 and 30 for the ImmunoPrep system, OptiLyse C, Uti‐Lyse, FACS Lysing Solution, and LF‐1000, respectively, but only 0.5 min for live whole blood. (ii) The forward and side scatter characteristics were affected by all of the lysis and fixation procedures, and this was most marked for LF‐1000‐Lyse and Flow. (iii) OptiLyse C gave preparations with poor forward and side scatter resolution due to the presence of residual red cell fragments. (iv) Lysis and fixation procedures did not affect the apparent expression of the CD3, CD45, or FMC7 antigens on normal or CLL samples, but gave highly variable results for the expression of the CD5, CD11b, κ, and λ antigens on the CLL samples. We conclude that lysis and fixation procedures can introduce different artefacts in the analysis of normal and leukaemic samples that are best avoided by analysing live whole blood. Cytometry (Comm. Clin. Cytometry) 38: 153–160, 1999. © 1999 Wiley‐Liss, Inc.