OBJECTIVES: Candida albicans is the most common fungal pathogen in humans, but other Candida species cause candidosis. Candida species display significant differences in their susceptibility to antimycotic drugs. Patients with symptomatic or erythematous oral lichen planus (OLP) commonly have Candida infection requiring correct identification of Candida species in order to initiate adequate antimycotic therapy. Therefore, conventional cytosmear and culture tests were compared with genetic diagnostics on oral rinse followed by agar culture and material collected by cytobrush from OLP patient mucosal lesion. METHODS: The genetic approach was validated on a reference panel of 60 well-defined unrelated fungal species. The study included 37 OLP patients. Oral candidosis (OC) was established based on clinical signs of OC and/or oral mucosal symptoms and at least one hypha in lesional cytosmear. Antimycotic treatment was initiated after OC diagnosis, and symptomatic treatment was initiated in no-candidosis situations. RESULTS: The composition of Candida species in oral rinse/culture test was different from that of lesional cytobrush sampling as more non-albicans species were detected by the latter. Unexpectedly, Candida dubliniensis was found to be overrepresented among patients with a history of antimycotic treatment indicating unintentional iatrogen selection. Of the 22 OLP patients receiving treatment, 27 % of these should have been offered alternative therapy based on the improved diagnostic approach. CONCLUSION: This study highlights the importance of lesional sampling in OLP patients with suspected OC. CLINICAL RELEVANCE: Correct fungal identification is critical in order to initiate adequate antimycotic therapy, thus minimizing iatrogen selection of non-albicans species.