An evaluation and a critique of prognostic parameters and criteria involved in the classification and clinical staging of epidermoid carcinoma of the vulva are presented, comparing the FIGO system and a proposed more definitive system. Biostatistical as well as clinical evidence for an improved system is presented. The proposed system is based on statistically valid data, has good patient distribution, and an orderly progression for therapy and prognosis. In addition, it is well suited for computer programming and for the comparison of patients and therapy, as well as for delineating the natural history of vulvar cancer.