OBJECTIVE: To determine the factors affecting survival following palliative large bowel resection for colorectal adenocarcinoma. PATIENTS AND METHOD: From the Colorectal Cancer Database of a single institution patients who had a palliative resection of a colorectal cancer from 1980 to 1993 inclusive were identified. Survival curves were constructed using the Kaplan-Meier method. Criteria studied were sex, age at operation, site of tumour, T, N and M status, tumour differentiation, involvement of tumour margins, tumour fixity and the presence or absence of peritoneal, liver or distant metastases. Multivariate analysis of factors was conducted using Cox proportional hazards analysis. RESULTS: Three hundred and seventy-seven patients (232 men, 145 women, median age 64 years) fitted the above criteria. Operative mortality was 5.6%. Crude 6 month survival rate was 71.1% and median survival 10.5 months. Significant factors affecting survival on univariate analysis were - Age (<75 vs. >75 years) (P=0.019); T status (T1/T2 vs. T3/T4) (P=0.039); nodal status (N0 vs. N1/N2) (P=0.0059); distant metastases (P=0.039) or liver metastases (P=0.0058); tumour differentiation (poor vs. moderate/well differentiated) (P < 0.001); involved tumour margins (P < 0.001). Multivariate analysis found the following factors significant: age (P=0.02), liver metastases (P=0.05), distant metastases (P=0.044), T status (P=0.042), nodal status (P=0.0063), tumour differentiation (P < 0.001) and involvement of tumour margins (P < 0.001). CONCLUSIONS: The data suggest that palliative resection of advanced colorectal carcinoma should be considered carefully in patients with advanced age, where distant metastases are present and in cases when primary tumours can not be completely resected. For the remaining patients, palliative resection may be accomplished with acceptable operative mortality and postoperative survival.