Abstract: Stabled horses commonly perform stereotypic patterns of weaving, where the horse shifts its weight from side to side often swinging its head. Ten warm-blood types, of which five were known to reliably weave, were housed in similar 12x12 ft wooden loose boxes in a single stable block surrounding a courtyard. Each horse was exposed to each of five stable designs. These were: the conventional front top-half of the door open only with a view of the stable courtyard (F); front half-door open and a similar half-door open at the back of the stable with a view to the surrounding fields (FB); back open only (B); front and one-side panel open with a view into the adjacent stable (FS); and front, back and both sides open (All4). During observation days, horses were brought in from the field at 0830 h, fed concentrate at 0930 h, fed haylage at 1005 h and turned out at 1600 h. Behaviour was recorded from 0900 to 1040 h, 1200 to 1300 h and 1500 to 1600 h. Weaving was most common prior to feeding in the morning and prior to putting out to pasture in the afternoon. There was a significant effect of stable design on weaving, with less weaving in the FS and All4 designs than the F treatment. There was also a significant effect of stable design on repetitive nodding, though in this case, FB, B, FS and All4 designs each reduced nodding compared with the F treatment. The effect of stable design can be explained in a number of ways. Firstly, it could be the novelty of the environmental change, though there was no evidence in this study of an increase in stereotypy with prolonged exposure to the new stable designs. Secondly, opening windows may increase opportunities for environmental interaction, and the expression of new activities may compete with stereotypic behaviour for the horse's time. Thirdly, the open windows may allow expression of specific activities such as environmental monitoring or social interaction that are denied by the conventional stable.