Clinical endometritis in dairy cows is defined as mucopurulent or purulent vulvar discharge 21 days or more after parturition. The diagnosis of clinical endometritis is commonly based on vaginal examination. Techniques to reduce the proportions of false negative findings have been described. This paper discusses a clinical approach to determine the proportion of false positive findings that might occur by vaginal inspection. The consequences of false positive findings in dairy practice are unnecessary or inadequate treatments. In research, incorrect diagnoses have an impact on the interpretation of studies on the diagnosis and treatment of clinical endometritis. The objective of the present study was to compare intrauterine bacteriology and endometrial cytology in cows diagnosed with clinical endometritis with findings obtained by vaginoscopy. Clinical endometritis was defined as mucopurulent or purulent vulvar discharge. On two commercial dairy farms, cows were examined 21 to 28 d postpartum. Uterine samples (n = 230) were collected from cows with clinical endometritis with the cytobrush technique to determine the proportion of polymorphonuclear neutrophils (PMN) and to culture smears for aerobic bacteria. Two threshold values for the proportion of PMN (5 and 18%) were chosen as possible indicators for an inflamed endometrium. Common uterine pathogens A. pyogenes and E. coli were found in 33.5 and 10.4% of the samples, respectively. With increasing vaginal discharge score, proportion of samples positive for A. pyogenes increased significantly. The proportion of cows exceeding the thresholds for PMN increased with vaginal discharge score and the presence of A. pyogenes. Considering only the presence of aerobic uterine pathogens and a proportion of PMN above the threshold values of 5 and 18% as indicative for endometritis, a proportion of 17.3 and 28.5%, respectively, of diagnoses by vaginoscopy were false positive.