BACKGROUND: Although laparoscopic resection of colorectal carcinoma improves post-operative recovery, long-term survival and disease control are the determining factors for its application. We aimed to test the null hypothesis that there was no difference in survival after laparoscopic and open resection for rectosigmoid cancer. METHODS: From Sept 21, 1993, to Oct 21, 2002, 403 patients with rectosigmoid carcinoma were randomised to receive either laparoscopic assisted (n=203) or conventional open (n=200) resection of the tumour. Survival and disease-free interval were the main endpoints. Patients were last followed-up in March, 2003. Perioperative data were recorded and direct cost of operation estimated. Data were analysed by intention to treat. FINDINGS: The demographic data of the two groups were similar. After curative resection, the probabilities of survival at 5 years of the laparoscopic and open resection groups were 76.1% (SE 3.7%) and 72.9% (4.0%) respectively. The probabilities of being disease free at 5 years were 75.3% (3.7%) and 78.3% (3.7%), respectively. The operative time of the laparoscopic group was significantly longer, whereas postoperative recovery was significantly better than for the open resection group, but these benefits were at the expense of higher direct cost. The distal margin, the number of lymph nodes found in the resected specimen, overall morbidity and operative mortality did not differ between groups. INTERPRETATION: Laparoscopic resection of rectosigmoid carcinoma does not jeopardise survival and disease control of patients. The justification for adoption of laparoscopic technique would depend on the perceived value of its effectiveness in improving short-term post-operative outcomes.