We have used a polymerase chain reaction-based assay measuring polymorphic (CA)n repeats, a class of simple sequence repeats, to assess the success of allogeneic canine marrow transplants. Results were compared with those obtained with karyotype analysis of dividing cells in recipients that were sex mismatched with their marrow donors. Twenty recipients were conditioned for transplantation of genotypically DLA-identical littermate marrow by 450 cGy of total-body irradiation. In 2 recipients, results could not be compared, since either only cytogenetic or dinucleotide (CA)n marker data existed. Both dogs had autologous marrow recovery. In 15 of the remaining 18 recipients, complete agreement was found between the results obtained with dinucleotide (CA)n markers, cytogenetic studies, and granulocyte changes after transplantation. Seven of the 15 showed eventual autologous recovery, 6 displayed mixtures of host and donor cells, and 2 showed donor-type hematopoiesis. Two of the 18 dogs showed mixed chimerism with (CA)n markers and autologous recovery by cytogenetics, findings that may be related to differences in cells analyzed by the two techniques--i.e., all nucleated cells by (CA)n markers versus dividing cells by cytogenetics. In one additional recipient, results of marrow cytogenetics, granulocyte changes, and (CA)n markers were consistent with a successful allograft, while peripheral blood cytogenetics suggested autologous recovery, possibly the result of erroneous blood sampling. Polymerase chain reaction-based testing for dinucleotide repeat (CA)n polymorphisms, originally developed for genetic mapping in the dog, is useful and reliable when compared with cytogenetic studies, in assessing the success of allogeneic marrow transplants in dogs.