Aim of this study was to compare uterine smears made using the Knudsen catheter, the cytology brush and a uterine culture swab with regard to diagnostic usefulness and the occurrence of neutrophils. Additionally correlation between culture results and the occurrence of neutrophils in uterine smears was investigated. Samples were collected from 340 mares, 81.5% of which were in estrus. Smears made using the cytology brush yielded more endometrial cells per high-power field than those made using the other two instruments (p<0.0001), and a larger proportion had PMNs compared with smears made using the uterine swab (p<0.0001). For smears made with the cytology brush, cultures of -hemolytic streptococci were more often (p=0.002) accompanied by PMNs than cultures of bacteria other than -hemolytic streptococci, and there was a positive correlation (r(s)=0.2 p=0.01) between the number of PMNs in smears and the number of colonies of -hemolytic streptococci. The cytology brush was superior to the other methods because it generated a larger proportion of diagnostic useful smears and the occurrence of PMNs in smears was significantly correlated with the occurrence of cultures of -hemolytic streptococci. Aim of this study was to compare uterine smears made using the Knudsen catheter, the cytology brush and a 20 uterine culture swab with regard to diagnostic usefulness and the occurrence of neutrophils. Additionally 21 correlation between culture results and the occurrence of neutrophils in uterine smears was investigated. 22Samples were collected from 340 mares, 81.5% of which were in estrus. Smears made using the cytology 23 brush yielded more endometrial cells per high-power field than those made using the other two instruments 24 (p < 0.0001), and a larger proportion had PMNs compared with smears made using the uterine swab (p < 25 0.0001). For smears made with the cytology brush, cultures of β-hemolytic streptococci were more often (p = 26 0.002) accompanied by PMNs than cultures of bacteria other than β-hemolytic streptococci, and there was a 27 positive correlation (r s = 0.2 p = 0.01) between the number of PMNs in smears and the number of colonies of 28 β-hemolytic streptococci. The cytology brush was superior to the other methods because it generated a larger 29 proportion of diagnostic useful smears and the occurrence of PMNs in smears was significantly correlated 30 with the occurrence of cultures of β-hemolytic streptococci. Introduction 38Bacterial endometritis is considered the most important reproductive disease in the mare and is associated 39 with considerable economic loss (Troedsson, 1997; Causey, 2006). However, in non-venereal bacterial 40 endometritis, risk factors such as compromised uterine defense mechanisms or poor perineal conformation, 41 rather than the bacterial organisms, constitute the primary problem (Causey, 2006). Persistent uterine 42 infections caused by β-hemolytic streptococci and E. coli from the vagina are a common sequel to 43 compromised uterine defense mechanisms (Causey,...