Therapeutic success and prognosis in RRP is determined by negative surgical margins. In order to minimize the proportion of positive surgical margins in the final histological evaluation, valid intraoperative control by means of frozen margin analysis is indispensable. We have developed and evaluated a new frozen margin procedure based on the Stanford method with transverse and sagittal cut directions. This technique facilitates comprehensive intraoperative evaluation of curved margin areas for the first time. Retrospective analysis of the results of the new frozen section technique revealed positive surgical margins in 2.7% of patients. The results obtained with this new technique were significantly superior to those obtained with two established techniques (10.3%, P < or =0.001; 17.2%, P < or =0.001). Our results demonstrate that the new frozen margin technique is clearly more sensitive for intraoperative detection of positive margins and thus leads to substantially higher rates of negative surgical margins.