Teat orifice hyperkeratosis, a commonly observed condition in dairy cows, has been considered a consequence of machine milking and the degree of hyperkeratosis may be increased by a poor milking system. A fully illustrated technique is described which uses a scoring system from 0 for a perfect orifice to 5 for an orifice significantly enlarged with extensively protruding fronds of teat duct keratin. A range of scores found in 25 commercial dairy herds is presented. The scores have been averaged for each cow and the markedly skewed distribution corrected by a square root transformation. This scoring procedure allows comparative measures of hyperkeratosis within and between herds. An 8-fold difference (0.17-1.31) in herd average score was found. Within all herds the score increased with lactational age of the animals and peaked, for any lactation, some 3-4 months post partum, declining as the animals dried off. There was no significant relationship between mean somatic cell count and degree of hyperkeratosis at the herd level. This implies that such chronic pathological changes are unlikely to be related to the level of intramammary infection. It appeared that some hyperkeratosis is an obvious and probably natural response to milking and occurs in a significant proportion of animals in all herds although often only to a slight degree. Much more hyperkeratosis may be a measure of the performance and management of the herd. The genetic influence is unknown. Higher yielding cows will score higher as they milk for longer, but generally high scores may reflect consistent and possibly considerable overmilking. Hyperkeratosis may be an indicator of the quality of management and show the level of attention being paid to the welfare of the herd.