182 samples of mammary tissue, 118 samples of axillary and axillary accessory lymph nodes as well as 526 bone samples and further routine tissue samples were collected from 27 bitches with malignant mammary tumours submitted for autopsy to the Institute of Veterinary Pathology of the Freie Universität Berlin in the years 1996 and 1997. These samples were investigated histologically and immunohistologically. Immunohistochemistry is carried out with the alkaline phosphatase anti-alkaline phosphatase (APAAP) method and the anti-cytokeratin antibody AE1. Among the breeds, terriers are affected in 19 % of the cases, dachshounds in 15 %. Mongrels are most frequently affected (40 %). The average age of the dogs is 11.3 years. All tissue samples of the mammary gland were routinely stained with H&E; and the tumours were classified. 27 main diagnoses stand against 40 diagnoses of malignant mammary tumours alltogether in 27 mammary chains. Of the main diagnoses there are nine adenocarcinomas, nine adenocarcinomas with hyperplasia of myoepithelium, four solid carcinomas, two complex carcinomas, two anaplastic carcinomas and one collusion tumour/carcinosarcoma. In nearly all cases, there is multiple tumour growth in the mammary chains. The number of malignant tumours grows steadily from cranial to caudal localisation. All 118 lymph node samples are stained with H&E; and immunohistochemically and diagnosed. The rate of metastasis in the lymph nodes is 40 % in the H&E; stain, whereas the immunostained lymph nodes are positive for metastasis in 54 % of cases. The immunohistochemical staining method is superior to the conventional method in so far as it reveals more micometastasis and it aids in determining the exact number and localisation of tumour cells in the lymph nodes. The localisation of tumour cells is orientated in the direction of the lymph flow through the lymph node. Thus 29 % of cases show tumour cells in the marginal sinus, near the marginal sinus or in the cortex, whereas 12 % of metastasis are located in the medullary zone. A lymph node with only one tumour cell is to be considered positive for metastasis, as the detection of a single tumour cell must be considered as a possible manifest metastasis. In 526 bone samples of 199 bones of 27 bitches, immunohistochemical staining reveals 14 % of the samples with metastasis or micrometastasis of a mammary tumour. The selection of the bone samples is done from predilection sites according to the literature. Ten humeri, five femur bones, nine pelvic bones and four lumbal vertebrae as well as one sternum show metastasis. 60 % of the positive reactions are found in the bone marrow, the spongiosa and the compacta with most of the reactions localised in the bone marrow. 40 % of positive reactions are localised directly around the bones. In the long bones, 46 % of positive reactions are in the diaphysis, 35 % in the metaphysis and 19 % in the epiphysis. The rate of metastasis related to the collective of the 27 bitches with malignant mammary tumours is for axillary lymph node metastasis 70.4 %, for bone metastasis 44.4 %. The bone metastases are presently not considered to be clinically relevant. The type of tumour is of major importance for metastasis of tumour cells in lymph nodes and in bones. Especially aggressive are solid carcinomas and anaplastic carcinomas, whereas adenocarcinomas with hyperplasia of myoepithelium show a lower rate of metastasis. An exact and thorough diagnosis of the tumour is necessary for the evaluation of the potential danger for metastasis. Particulary negative in the prognosis in histologic tumour assessment are thereby: lymph- and hemangiosis carcinomatosa, solid and anaplastic tumour growth, the presence of tumour giant cells, polymorphous cells, no cystic component of the tumour, abundant stroma, tumours interspersed regulary and abundantly with vessels, multiple nucleoli, necrosis and hemorrhage or inflammatory reactions in and around the tumour. A mammary tumour located in the caudal abdominal or inguinal mammary complex has the potential to metastasize to the axillary lymph node. This, the high rate of metastasis to the axillary lymph nodes and the fact, that in all mammectomised bitches metastases to the axillary lymph nodes were found, presently pose a question mark to the usual surgical treatment of the mammary tumours, that is the exstirpation of the mammary chain with the inguinal lymph nodes. In metastasis of tumour cells via the blood to the bones, the vertebral vein plexus that contains veins without valves is to be noted. This makes it possible to explain high rates of metastasis to the bones, that are apparenttly considered too low in veterinary literature. Furthermore, a direct spread of tumour cells to the bones of the axial sceleton would be possible without the tumour cells passing the lungs. Organs of nearly three quarters of the bitches are affected by metastasis following a primary tumour of the mammary gland. Most cases show a combination of lymph node metastasis and metastasis to other organs. If metastasis into organs essential for life functions are present, a surgical therapy is not indicated.