ABSTRACTThe FIGO classification for the clinical staging of carcinoma of the cervix does not adequately express primary tumour extent or volume, and contains anomalies which probably lead to a wide variation in therapeutic results between centres (particularly in Stages II and III). A box diagram method of recording primary tumour extent is proposed and was applied to a retrospective study of 191 women aged 35 years and under with invasive carcinoma of the cervix. Patients who scored one or two boxes (Stages IB, IIA and IIB with unilateral parametrial infiltration) have a similar actuarial 5 years survival rate: 70% and 69% respectively. Those with 3 boxes (Stage IIA and B, bilateral Stage IIB and unilateral IIIB) and 4 boxes or more (Stages IIA and B, bilateral IIIB) involved had a substantially worse survival: 44% and 22% respectively. No patient, with more than five boxes involved was cured. The further study of this system, its potential in clinical research, and as part of a possible revised staging system are discussed.