5,6-dimethylxanthenone-4-acetic acid, a novel antivascular anticancer drug, has completed Phase I clinical trial. Its actions in mice include tumour necrosis factor induction, serotonin release, tumour blood flow inhibition, and the induction of tumour haemorrhagic necrosis and regression. We have used mice with a targeted disruption of the tumour necrosis factor receptor-1 gene as recipients for the colon 38 carcinoma to determine the role of tumour necrosis factor signalling in the action of 5,6-dimethylxanthenone-4-acetic acid. The pharmacokinetics of 5,6-dimethylxanthenone-4-acetic acid, as well as the degree of induced plasma and tissue tumour necrosis factor, were similar in tumour necrosis factor receptor-1 7/7 and wild-type mice. However, the maximum tolerated dose of 5,6-dimethylxanthenone-4-acetic acid was considerably higher in tumour necrosis factor receptor-1 7/7 mice (4100 mg kg 71 ) than in wild-type mice (27.5 mg kg 71 ). The antitumour activity of 5,6-dimethylxanthenone-4-acetic acid (25 mg kg 71 ) was strongly attenuated in tumour necrosis factor receptor-1 7/7 mice. However, the reduced toxicity in tumour necrosis factor receptor-1 7/7 mice allowed the demonstration that at a higher dose (50 mg kg 71 ), 5,6-dimethylxanthenone-4-acetic acid was curative and comparable in effect to that of a lower dose (25 mg kg 71 ) in wild-type mice. The 5,6-dimethylxanthenone-4-acetic acid -induced rise in plasma 5-hydroxyindoleacetic acid, used to reflect serotonin production in a vascular response, was larger in colon 38 tumour bearing than in non-tumour bearing tumour necrosis factor receptor-1 7/7 mice, but in each case the response was smaller than the corresponding response in wild-type mice. The results suggest an important role for tumour necrosis factor in mediating both the host toxicity and antitumour activity of 5,6-dimethylxanthenone-4-acetic acid, but also suggest that tumour necrosis factor can be replaced by other vasoactive factors in its antitumour action, an observation of relevance to current clinical studies.