We have recently developed a computer-assisted olfactory test to measure detection thresholds for five pure odorants. The reference group consisted of 30 subjects without olfactory complaints. Statistical analysis was carried out to identify a statistical criterion for determining olfactory deficiencies. This criterion was applied to 54 subjects suspected to have an olfactory deficiency, either on the basis of their subjective complaints or on clinical examinations (e.g. scanner radiography, endoscopic investigations, rhinomanometric measurements). Nine aetiological groups were screened: trauma, nasal polyposis, nasal obstruction, allergic rhinitis, post-influenza, post-anaesthesia, endocrine dysfunction, hereditary hyposmia, and subjective olfactory loss without a clear aetiology. In each group, this method allowed us to discriminate between deficient and non-deficient patients, and the olfactory deficit could be quantified. This rapid procedure was well-accepted by all subjects and gave reproducible quantitative results. It can provide useful information about the relationship between olfactory acuity and a given aetiologic category.