INTRODUCTION: Neoadjuvant radiochemotherapy (neoRT/CT) in locally advanced rectal cancer requires an exact initial determination of the depth of the cancerous infiltration (T-status) and of locoregional lymph node metastasis (N-status). For staging and restaging, contrast-enhanced computed tomography (CT) is usually used. In specialised centers, the endorectal ultrasound (rES) may be preferred. METHODS: Between January 1998 and May 2001, the T- and N-status of 102 patients with adenocarcinoma of the rectum (> or =T3 or N+) was determined prospectively by rES and CT (group I: n=61 without neo-RT/CT, examined once; group II: n=41 examined before and after neoRT/CT). All diagnostic findings were compared using the (y)pTNM-classification. RESULTS: In the patients from group I, the depth of infiltration (uT) was predicted correctly by rES in 75% and by CT in 48% of cases; the carcinomas were understaged in 10% and 41% of cases and overstaged in 15% and 11%, respectively. According to the histopathological findings, the N-status was determined correctly by rES and CT in 75% and 57% of cases, understaging occurred in 8% and 30% and overstaging in 17% and 13%, respectively. In cases in which both methods resulted in identical T- (uT+ctT) or N-staging (uN+ctN), the accuracy increased to 82% and 80%, respectively. In patients from group II, after neoRT/CT rES and CT allowed the exact prediction of the yuT-stage in 66% and 51%, respectively. Only 2% were understaged by rES (understaging by CT: 22%). Overstaging occurred in 32% and 27% by rES and CT, respectively. The N-status determined by rES and CT was in accordance with the histopathological findings in 68% and 76%of cases, respectively. Understaging occurred in 20% and 17%,overstaging in 12% and 7%, respectively. Again identical staging results in both rES and CT increased the accuracy of the T- (yuT+yctT) or N- (yuN+yctN) classification to 90% and 83%, respectively. In group II, downsizing of the tumor by more than one T-stage was correctly assessed by rES results in 15/20 cases (75%). A complete remission of initial uT3-carcinoma was diagnosed correctly in only two of eight ypT0-cases. In contrast, CT demonstrated a remission of disease in all cases but was unable to predict the extent of tumour reduction. A remission of lymph node metastasis was accurately shown by rES in 17/19 cases (90%) and by CT in 10/12 cases (83%). CONCLUSION: The staging of pretherapeutic, locoregional T- and N-status by rES is superior to that by CT (T-status: P=0.0164, N-status: P=0.0035). At restaging, rES offers higher accuracy in the detection of residual tumour infiltration (but not significantly to CT, yT-status: P=0.0833, yN-status: P=0.7962) and assessment of local remission. Therefore rES should be the method of choice in staging to avoid overtreatment in neoadjuvant settings.After neoRT/CT, the predictive efficacy of the rES for the downsizing/-staging of rectal cancer must be evaluated on greater numbers of patients receiving standardised diagnostic procedures and therapy.