Early embryonic death (EED) is a potential cause of infertility in mares. Causes include embryo related chromosomal defects or disturbances in the oviductal and uterine environment, but often the cause remains unknown. In humans, most of the endometrial lymphocytes are natural killer (NK) cells (CD56-positive lymphocytes), and their phenotypes, surface receptors, and cytokine production have been associated with early embryonic death and endometrial angiogenesis. (Fuchinoue et al. J Obstet Gynaecol Res. 2016;42:1541-1552). The aims of this study were to 1) describe the relationship between endometrial biopsy grade and future fertility in mares with repeated EED, and 2) to identify and localize equine natural killer cells in the endometrium of mares with repeated EED. The study included 32 formalin-fixed paraffin-embedded endometrial biopsies of mares with a history of two or more EEDs. Immunohistochemistry for the identification of NK-cells was done on endometrial biopsies, using polyclonal NK cell surface receptor antibodies NK2GD and CD56 (Thermo Fisher Scientific, MA, USA) according to manufacturer instructions. Mouse brain and spleen samples functioned as positive controls. Information on biopsies and post-biopsy foaling were collected and analysed with Fisher'sexact test to evaluate dependence between variables (age group, biopsy grade, later abortions, previous parturitions, number of EED, and live foals). A p-value of < 0.05 was considered significant. The biopsies were graded according to Kenney & Doig (In: Morrow, DA, ed. Current Therapy in Theriogenology. 1986; 723–729). Of all mares, 3.1% were grade IIA, 43.8% grade II, 53.1% grade IIB. In total, 43.8%were younger than 13 years and 56.3% 13 years or older, 71.9% had had two EEDs and 28.1% more than two. In total 71.9% of the mares had foaled previously, 68.8% were bred again of which 59.1% foaled a live foal. The age group (< 13 vs. ≥ 13), biopsy grading (grade I-II vs. IIB-III) or later abortions (abortion at over 50 days of pregnancy) were found not to have a significant effect on the later live foal rate. Previous foalings and the number of EEDs (< 3 vs. ≥ 3) seemed to affect the foaling result but failed to have statistical significance in this small study population. Natural killer cells were not identified in the endometrial samples with the selected antibodies, suggesting potential species differences in surface receptors. In conclusion, repeated EED was not related to lower future foaling rates, implying mainly incidental causes of EED in this population.