The cross-reactivity of a group of monoclonal antibodies (MABs) generated against human cytokeratins (CKs) was investigated in mouse tissues. Formalin-fixed and paraffin-embedded sections of lung, stomach, small and large intestine, liver, and kidney were immunostained with MABs after epitope retrieval with enzyme digestion. AE1/AE3, a "cocktail" of two MABs that recognizes basic and acidic CKs, 5D3 MAB to low molecular weight CKs (8, 18, and 19), and monospecific MABs to CK 7 and 20 were tested. Additionally, CK 17 and 34betaE12 MABs to high molecular weight CKs were evaluated in the same organs and in sections from skin and preputial glands. We employed the new universal animal system (ARK) as the detection system. The results showed intense reactivity for the first group of antibodies used, with topographic distribution similar to that in human tissues, with the exception of CK 7 in lung parenchyma, which displayed reactivity only in type II pneumocytes, with negativity of adjacent bronchial epithelium. Also of note was the lack of reaction of liver hepatocytes and renal tubular cells to AE1/AE3 and 5D3 MABs. Regarding the second group of antibodies, no reaction was obtained for CK 17 in the tissues tested. On the contrary, 34betaE12 MAB yielded intense reactivity in cells of epidermis and hair follicles. Compared to other detection systems used previously in this animal, ARK produced a well-defined reactivity at the cellular level without any background. We conclude that a useful panel of anti-CK antibodies commonly used in human pathology can be applied successfully to mouse tissues after enzyme digestion, leading to a more accurate definition of cellular populations in this laboratory animal.