Emissions from dairy cow housing are an important term in the UK inventory of ammonia emissions from agriculture: even by current estimates, they contribute, on their own, 9% of total national emissions. When constructing such inventories, it has so far always been assumed that emissions arise only over the winter when the cows are housed full rime, or almost so, and that the emissions from the remainder of the year are zero. This study aimed to test that assumption by suitable experiments. A preliminary survey showed that ther-e are wide variations in summer housing practice from dairy farm to dairy farm, and the practice of not housing the cows at all during the summer is by no means universal. Therefore, the dairy cow house chosen for measuring ammonia emissions under summer and winter conditions was one at which the summer practice (which was of a very common type) resulted in cows spending an average of 4 h in every 24 inside the house. Using ＇Ferm tube＇ passive ammonia flux samplers, the average summer ammonia emission rare determined from four separate experiments on that howe was not significantly different, at the 5% level of probability, from the average winter ammonia emission rate determined from four separate experiments. The main implication of the above findings is that the assumption mane hitherto, in both the UK and the Netherlands, when building ammonia inventories, that summer emissions fr-om dairy cow houses are negligible, is untenable. Further measurements on summer emission rates with other common summer housing practices are needed, but meanwhile a first (perhaps pessimistic) estimate can be made of the necessary adjustment in overall ammonia emissions, by doubling the present term in the national inventory for winter emissions. The result of so doing is to increase the UK total ammonia emissions from agriculture by 9% (i.e. by 18 kr NH3-N).