Octreotide, a somatostatin analogue, has been shown to inhibit the growth of gastrointestinal cancers in vitro and in vivo. To assess the anti-tumour effect of octreotide, we performed a randomised trial comparing octreotide with best supportive care in advanced gastrointestinal cancer patients refractory to chemotherapy. A total of 107 patients with advanced gastrointestinal cancer refractory to chemotherapy were randomised to receive octreotide at the dose of 200 micrograms three times a day for 5 days a week, or the best supportive care only. The primary outcome variable was the survival duration. Response rate was an outcome variable of secondary importance. Fifty-five patients (15 stomach, 16 pancreas, 24 colon-rectum) received octreotide, while 52 (14 stomach, 16 pancreas, 22 colon-rectum) received the best supportive care. Patients treated with octreotide had a significant advantage in duration of survival with a median survival time of 20 weeks vs 11 in the control group (P < 0.0001). This advantage was present also considering the survival data for each tumour group. Twenty-five patients (45%) given octreotide showed stable disease vs only eight (15%) in the control group (P < 0.001). In conclusion, octreotide therapy seems to confer a survival benefit in advanced gastrointestinal cancer patients refractory to chemotherapy. Additional studies will be needed to confirm these results and to clarify other questions about dose and schedule of octreotide.