This study aimed to evaluate the localised effects of intrauterine ozone therapy on endometrial recovery in mares with endometritis. Our investigation assessed changes in gene expression profiles of anti‐inflammatory (IL‐1RA and IL‐10), proinflammatory (IL‐R1B3i and TNFα) and pleiotropic (IL‐6) cytokines, along with detailed histological measurements of epithelial and endometrial thickness and the glandular area ratio. Twenty mares were assigned to a 2 × 2 factorial design based on endometritis diagnosis and treatment (control or 42 μg/mL ozone insufflation), resulting in four groups: NC (negative for endometritis/control), NO (negative/ozone), PC (positive/control) and PO (positive/ozone). Oestrus was induced with 2 mg of oestradiol benzoate on Days −1, 1 and 3, plus 1 mg on Day 5. Day 0 marked the initial uterine treatment, followed by insufflations on Days 1 and 2 with O3 (ozone) or O2 (control). Uterine biopsies were taken before treatment on Day 0 and Day 6 for histological analysis and gene expression assessment. Data were analysed using a statistical model that included endometritis status, treatment type, biopsy times (D0 and D6) and their interactions, analysed with Proc Glimmix. Regardless of treatment or endometritis status, significant biopsy effects (p < 0.01) indicated increased epithelial height and endometrial thickness in Day 6 samples. Analysis of IL‐1 and TNFα revealed a significant interaction (p < 0.05) among endometritis, treatment and biopsy, with higher IL‐1B3i expression on Day 6 in the PC group. The treatment effect (p < 0.04) showed a higher frequency (p < 0.01) of animals with positive modulation in the PC group (66.7%) versus the PO group (0.0%). An interaction effect (p = 0.08) between endometritis and treatment resulted from higher IL‐1RA expression on Day 6 in the PC group compared to the PO group. Biopsy effect was significant for IL‐10 (p < 0.01), indicating higher values in the second sample associated with tissue repair. In the short‐term evaluation, ozone therapy did not influence endometrial morphology and may modulate cytokine expression, specifically the reduction in IL‐1 and TNFα levels. Therefore, this therapy appears to be a safe and potentially effective treatment for modulating the inflammatory response in mares with endometritis.