Experiments were performed to identify subpopulations of lymphocytes in lymphoid tissues of dogs. Anti-canine thymocyte serum (ATS) and anti-canine brain-associated T-cell antigen (ABAT) were obtained by immunization of rabbits with canine thymocytes and canine brain tissue homogenate, respectively. The ATS and ABAT antisera were absorbed using canine RBC, liver powder, and bone marrow cells. Part of the ABAT antiserum was further absorbed using canine thymocytes (ABAT-T antiserum). The ATS, ABAT, ABAT-T, and antiasialo GM1 antisera were used for identification of lymphocyte subpopulations by means of indirect immunofluorescent antibody (FA) techniques. For bone marrow-derived lymphocytes, surface immunoglobulin (SIg) was measured by direct FA using fluorescein isothiocyanate-conjugated rabbit anti-dog immunoglobulin G antiserum. The specificities of the various antisera were proven by treatment of the lymphocytes by means of nylon wool columns or rosette-forming techniques followed by treatment with antisera and rabbit complement. Results indicated that ATS reacted with T-lymphocytes, and ABAT antiserum reacted with T-lymphocytes, with ABAT-T-positive cells, and with antiasialo GM1-positive cells The ABAT-T and asialo GM1 antisera reacted with the same cell populations, but did not react with ATS-positive cells and ABAT-positive cells. Lymphocytes from 21 clinically normal mixed-breed dogs and from 2 Beagle dogs were identified by various antisera. For mixed-breed dogs, the positive percentages of peripheral blood lymphocytes with respect to SIg, ATS, ABAT, and ABAT-T were 36.3 +/- 3.2%, 52.7 +/- 3.5%, 56.1 +/- 3.6%, and 2.4 +/- 0.3%, respectively; in the spleens, the positive percentages were 44.8 +/- 3.8%, 39.6 +/- 2.8%, 42.6 +/- 2.6%, and 1.9 +/- 0.5%, respectively; and in the lymph nodes, 39.0 +/- 2.7%, 49.4 +/- 3.6%, 52.1 +/- 3.9%, and 1.6 +/- 0.5%, respectively.(ABSTRACT TRUNCATED AT 250 WORDS)