The aim of this prospective split sample study was to evaluate the applicability of liquid-based cytology (LBC) of oral brush biopsies for detection of oral cancer. Two different preparation techniques were investigated: the conventional transfer procedure to glass slides and the LBC preparation method. The obtainments of epithelial cells were performed five times with a nylon brush and transferred onto five glass slides. Additionally, the brushes, which were normally discarded, were stored in a fixative solution. Conventional slides and respective thin layers from a total of 113 cases were reviewed with both techniques. Thin layers showed excellent morphology on a clear background, which allowed an accurate diagnosis. In contrast, the conventional glass slides showed significantly more blood contamination and cell overlapping. The sensitivity of conventional cytological diagnosis was 96.3%, the specificity archived 90.6%, the positive predictive value was 96.3% and the negative predictive value scored 90.6%. The sensitivity of the cytological diagnosis using thin layers archived 97.5%, the specificity was 68.8%, the positive predictive value revealed 88.76% and negative predictive value was 91.7%. Our findings indicate that in oral cytology, LBC may replace other types of wet-fixed preparations using the full amount of collected cells, resulting in enhanced specimen quality archiving comparable values of diagnostic accuracy. LBC facilitates the cell collection due to simpler handling and less transfer errors by dentists and may improve the overall diagnostic accuracy of oral brush biopsies in future.