In order to study the effects of sheep teat disorders on the protection of the mammary gland, we used a Mannheimia haemolytica isolate, which did not cause clinical mastitis when deposited into intact teats. In the first experiment, this was deposited into the duct of teats with orf (Group A, n=5) or papilloma (Group B, n=3). In the second, teats were chapped and then, the organism was deposited into the duct (Group C, n=7) or on the skin (Group D, n=4). Ewes with healthy teats were controls (Group E, deposition into duct, n=5; Group F, deposition on skin, n=2). The ewes in Groups A, B or C developed clinical mastitis 5 h later, whilst the ewes in Group D developed it 2 d later; no control ewe developed clinical mastitis. In ewes with teat lesions, the organism was isolated from secretion samples and the California Mastitis Test became positive 5 h after challenge; neutrophils and lymphocytes were seen in Giemsa-stained secretion films from Group A or B ewes, whilst macrophages, neutrophils and lymphocytes in films from Group C or D ewes; neutrophils were predominating in films from Group E or F ewes. Inside the teats of Group A, B, C or D ewes, folds, hyperaemia and mucosal thickness were seen; histologically, subepithelial leucocytic infiltration was seen. In Group A or B ewes, no evidence of lymphoid tissue at the teat duct-cistern border was found. In Group C or D ewes, intense erosion and ulceration of the teat skin and conspicuous lymphoid tissue at the teat duct-cistern border, were evident; lesions characteristic of haemorrhagic mastitis were in the mammary parenchyma. In control ewes, subepithelial leucocytic infiltration in the teat duct and lymphoid tissue as above, were evident. We postulate that teat lesions can be predisposing factor to mastitis, by adversely affecting defences and speeding the process of infection and making it more severe.