OBJECTIVE We retrospectively analysed our experience of contact therapy alone and/or combined with interstitial brachytherapy as exclusive treatment of low lying rectal tumours. METHODS From 1971 to 2001, 124 patients (103 adenocarcinomas, 21 villous tumours) were treated by contact therapy alone or combined with interstitial brachytherapy. All patients were staged according to the Dijon classification. The average size of the lesions was 2.4 cm (max 7 cm), clinical aspect was polypoïd in 75% of the cases, flat in 17%. Sixty four patients received contact therapy in three fractions and 44 patients received four fractions, for an average delivered dose of 95 Gy. Interstitial brachytherapy boost delivered 24 Gy on a reference isodose of 55 cGy/h in 10 patients. RESULTS The local control was 83% for T1 and 38% for T2 tumours (p=0.004). For mobile tumours, the local control rate is 76%, significantly higher than for tumours with impaired mobility (55%, P=0.03). Thirty-nine patients experienced a local failure (31%). For patients amenable to surgery, a Miles procedure was performed in 25 patients. Ultimate local control rate is 93% for T1, 69% for T2 (P<0.05), 15 patients failed despite treatment for local recurrence (15%). No significant differences were observed in a comparison of adenocarcinoma and villous tumours according to initial and ultimate local control. The mean disease free survival rate for the whole population is 66 months. The 5-year disease free survival for T1a and T1b is, respectively, 82 and 78%, 40 and 25% for T2a and T2b, respectively. The overall 5-year survival for the whole group is 62.4%. At the end of the treatment, 75% of the patients described a very good sphincter function. No deleterious effect on continence was reported during the follow-up. CONCLUSIONS The control rate for T1 rectal cancer treated with contact therapy with or without brachytherapy is comparable to surgical series. The sphincter was preserved in 80% of the patients. Radiotherapy remains an efficient and cheap alternative to surgery, mainly for old and fragile patients, or refusing colostomy. The results of these approaches for tumors larger than 3 cm (T2) are not satisfactory. For patients not amenable to surgery, external beam radiation therapy and/or combined modality with chemoradiation should be discussed to increase the loco-regional control rate. A careful selection of patients based on rectal examination and trans-rectal ultrasound could select more accurately patients amenable to such an approach.