Oestrous suppression by intrauterine devices (IUDs) is caused by prolongation of luteal function, but the biological mechanism is unknown. The aim of the study was to investigate mechanisms which could explain the action of IUDs. Thirty mares were age-matched and either inseminated (AI, n = 15) or fitted with an IUD (IUD, n = 15) and subsequently divided into four groups: AI-P, pregnant (n = 8); AI-N, non-pregnant (n = 7); IUD-P, prolonged luteal phase (n = 7); and IUD-N, normal luteal phase (n = 8). The median ages were 5.5 and 7 years in AI-P and IUD-P groups and 14 and 11 years in AI-N and IUD-N groups, respectively. On Day 15 after ovulation, an endometrial biopsy was obtained to study histomorphological and immunohistochemical expression patterns of uterine proteins (uteroferrin, UF; uterocalin, UC; uteroglobin, UG), oestrogen and progesterone receptors (ER, PR), proliferation marker Ki-67 and content of inflammatory cells. Expression of UF was higher in IUD mares; the difference between pregnant and IUD-P mares was significant. Mares exhibiting a prolonged luteal phase (AI-P, IUD-P) showed only mild angiosclerosis and lower expression of both ER and PR than mares with a normal luteal phase (AI-N, IUD-N). No significant differences were detected in the numbers of inflammatory cells, with the exception of macrophages, which were more numerous in AI-P than AI-N mares. Although inflammatory cells were not detected in IUD mares, increased UF levels may indicate chronic inflammation. Young age and normality of the endometrial blood vessels may improve the efficacy of IUDs.