Allergic rhinitis (AR) is the most common allergic disease. The Allergic Rhinitis and its Impact on Asthma (ARIA) guidelines classify AR according to its duration and severity and suggest recommended treatments, but there is evidence that these guidelines are insufficiently followed. Considering the validity of histopathological data, physicians are more likely to be persuaded by such information on AR. Thus, we attempted to define the severity of AR by nasal cytology on the basis of the ARIA classification. We examined 64 patients with AR caused by sensitization to grass pollen. We clinically defined AR according to the ARIA classification and performed nasal cytology by Rhino-probe sampling, staining and reading by optical microscopic observation. Clinically, 22 (34.4%), 21 (32.8%), 10 (15.6%), and 11 (17.2%) patients had mild intermittent, moderate-to-severe intermittent, mild persistent, and moderate-to-severe persistent AR, respectively. Nasal cytology detected neutrophils in 49 patients, eosinophils in 41 patients, mast cells in 21 patients, and lymphocytes or plasma cells in 28 patients. The patients with moderate-to-severe AR had significantly more mast cells and lymphocytes/ plasma cells than those with mild AR. Our findings demonstrate that the ARIA classification of AR severity is associated with different cell counts in nasal cytology; especially, moderate-to-severe AR shows significantly increased counts of mast cells and lymphocyte or plasma cells. The ease of performing nasal cytology ensures is feasibility as an office AR diagnostic procedure for primary care physicians, able to indicate when anti-inflammatory treatments, such as intranasal corticosteroids and subcutaneous or sublingual allergen immunotherapy, are needed.