Various methods of separation and identification of major equine leukocyte populations and subpopulations were used. The purity of T and B lymphocytes separated in Sephadex anti-equine F(ab')2 columns was 87% to 99% and 83% of 97%, respectively. The purity of T lymphocytes separated in nylon-wool columns was 89% to 98%. Preparations of B lymphocytes separated in glass-bead columns were 68% to 79% pure. The presence (or absence) of surface immunoglobulin by immunofluorescence was the most consistent and reliable method for the identification of B or T lymphocytes, respectively. However, the erythrocyte-antibody-complement-rosette method for the identification of B cells and the erythrocyte-rosette method for the identification of T cells were not suitable. Monocytes were separated by the adherence method, and the purity, as identified by the latex particle ingestion procedure, was 70% to 78%. Electron microscopy of monocytes stained by peroxidase activity did not identify these cells. The purity of neutrophils obtained by the Ficoll-Hypaque separation method was 95% to 97%. The merits and usefulness of these methods were discussed.