Abstract Study question Does oral ofloxacin 2x200 mg for 10 days combined with vaginal metronidazole 1x500 mg for 10 days reduce the activity of chronic endometritis? Summary answer The investigated antibiotics may be beneficial in some cases of chronic endometritis, except in cases of concomitant endometrial polyp, where they adversely affect treatment outcome. What is known already Chronic endometritis (CE) is characterized by the presence of atypical plasma cell infiltrates (CD138 positive) in the endometrial stroma. CE may adversely affects fertility by reducing endometrial receptivity, impairing decidualization and uterine contractility, thus increasing the risk of recurrent pregnancy loss and implantation failure. Some idiopathic infertility cases are attributed to CE. The diagnostics of CE is challenging because the clinical evaluation is nonspecific and the immunohistochemical criteria are not consistent between research centers. Literature data suggested that successful treatment of CE with antibiotics could improve live birth rates, but a preferred treatment regimen has not been identified. Study design, size, duration A prospective cohort study included women undergoing office hysteroscopy (OH) at a tertiary referral center in 2021-2023. Women diagnosed with CE were assigned, based on their preference, to the intervention arm of ofloxacin 2x200 mg orally for 10 days + metronidazole 500 mg vaginally for 10 days or to the control arm (no antibiotics). 100 women underwent repeated hysteroscopy after 3 cycles, of which 77 received antibiotics and 23 were controls. Participants/materials, setting, methods Women aged 18-45 who met eligibility criteria for OH were included. Karl Storz 5.0 mm Bettocchi® operating sheath with a 2.9 mm 30° telescope and a 5Fr working channel was used. Endometrial lesions or biopsy specimens were collected with grasping forceps, blunt scissors or bipolar needle electrode. Saline was used as the dilating medium. Tissue samples were subjected to histopathological examination. Plasma cells were detected by immunohistochemical staining with Monoclonal Mouse Anti-Human CD138 antibodies. Main results and the role of chance The most common findings identified at hysteroscopy were endometrial polyps (43%), polypoid endometrium (38%), micropolyps (23%), focal (27%) and diffuse hyperemia (18%) and cesarean scar defect (14%). CE was defined as the presence of > 2 plasma cells/mm2. The Mann-Whitney U test showed a significantly higher plasma cell count (PCC) in the presence of a cesarean scar defect than in other cases (23 vs. 18/mm2, p = 0.031) and a tendency towards a lower PCC in micropolyps (9 vs. 21/mm2, p = 0.063). The correlation of hysteroscopic findings with CE activity was weaker than expected. Using repeated measures ANOVA, no difference in PCC between primary and follow-up hysteroscopy was found in either arm (p = 0.8), while post-hoc tests showed that the mean PCC decreased from 12 to 5/mm2 (p = 0.02) in the intervention arm, and from 14 to 6 (p = 0.23) in the control arm. On follow-up hysteroscopy, polyp remission was observed significantly more often in the control arm than in the intervention arm (61% vs. 27%, pChi2=0.02), in which de novo polyp was observed in 4%. In polypoid endometrium, the remission rate was similar in both arms (19% vs. 22%), and the rate of de novo lesions was significantly lower in interventional arm (16% vs. 39%, pChi2=0.035). Limitations, reasons for caution The limitations include the small sample size, the CE definition based on the number of plasma cells calculated using the Receiver Operating Characteristic curve for the purposes of the study, and the empirical selection of antibiotic therapy. Wider implications of the findings Due to the weak correlation of hysteroscopic findings with CE activity, the diagnostics and treatment of CE based on clinical evaluation of the uterine cavity is not justified. Poorly selected antibiotic therapy may adversely affect the outcomes of treatment. Trial registration number 1072.6120.322.2020