A study was carried out on 10 barren mares to evaluate comparative efficacy of 3 diagnostic techniques of endometritis, viz., endometrial biopsy (EB), cytobrush (CB) and low-volume uterine lavage (LVL). The samples collected were evaluated cytologically and bacteriologically. CB yielded significantly more PMNs per high power field (HPF) (57.33±0.74) than EB (35.50±1.54) and LVL (14.35±1.05). In LVL, the proportion of PMNs to uterine epithelial cells (1.20%) was higher than in CB (0.86%) and EB (0.48%). EB yielded significantly more intact cells (90%) than CB (60%) and LVL (42%). Distorted cells were significantly more in smears made from LVL (48%) than from EB and CB (9% each). Fragmented cells were significantly more in CB (31%) than EB (1%) and LVL (10%) samples. Inflammatory cells, other than PMNs, were detected in CB and EB, and vaginal epithelial cells only in LVL. According to criterion II (≥1% PMNs/300 cells), endometritis was diagnosed in 30% mares by EB and CB and 40% by LVL. None of the mares was positive as per criterion I (≥1 PMNs/HPF). Positive microbial growth was obtained in 7, 6 and 8 mares using EB, CB and LVL. All the cytologically positive mares were also positive for microbial cultures. Whereas, 30% positive cultures from CB and 40% from EB and LVL had negative cytology. The most common pathogenic bacteria cultured from CB and EB were β-hemolytic Streptococci and E. coli, whereas in LVL, E. coli was isolated more frequently. Staphylococcus spp., Pseudomonas spp. and Yeast were also isolated mostly in combination with other microbes. The combined infections were seen only in EB and LVL samples. Cloudy or mucoid effluxes of LVL were mostly associated with isolation of E. coli and β-hemolytic Streptococcus (5/10). The recovery of Streptococcus was associated with a rise in uterine pH (8.0), but not the E. coli (pH 6.5). In all three methods, number of samples culturally positive for β-hemolytic Streptococcus were always associated with a positive cytology, but not the E. coli.