The aim of the present study was to evaluate the definition of oral leukoplakia, proposed by the WHO in 2005 and taking into account a previously reported classification and staging system, including the use of a Certainty factor of four levels with which the diagnosis of leukoplakia can be established. In the period 1997-2012 a hospital-based population of 275 consecutive patients with a provisional diagnosis of oral leukoplakia has been examined. In only 176 patients of these 275 patients a firm diagnosis of leukoplakia has been established based on strict clinicopathological criteria. The 176 patients have subsequently been staged using a classification and staging system based on size and histopathologic features. For use in epidemiological studies it seems acceptable to accept a diagnosis of leukoplakia based on a single oral examination (Certainty level 1). For studies on management and malignant transformation rate the recommendation is made to include the requirement of histopathologic examination of an incisional or excisional biopsy, representing Certainty level 3 and 4, respectively. This recommendation results in the following definition of oral leukoplakia: 'A predominantly white lesion or plaque of questionable behaviour having excluded, clinically and histopathologically, any other definable white disease or disorder'. Furthermore, we recommend the use of strict diagnostic criteria for predominantly white lesions for which a causative factor has been identified, e.g. smokers' lesion, frictional lesion and dental restoration associated lesion.