Aims/hypothesis. To describe a unifying hypothesis of the relation between insulin resistance and inflammatory response in the development of diabetes. Methods. Review of the literature and authors' research. Results. Infection and injury activate the immune system and bring about widespread metabolic changes which disavantatge and destroy the invading organism and facilitate repair of damaged tissue. Tumour necrosis factor-α is involved in inflammatory events and fight against infection. No study has extensively investigated its numerous metabolic effects. From induction of hyperlipidaemia to regulation of intracellular insulin signalling, TNF-α has been even associated with nutrient-sensing pathways. Certain TNF- α gene polymorphisms (linked to a high transcription rate of TNF-α), and the plasma concentrations of the TNF-α soluble receptor are simultaneously associated with insulin resistance, body fat, and with mortality after chronic infections. Thus, the TNF system seems to be designed for an effective fight against infection and for providing survival advantages during periods of food shortage. By inducing muscle insulin resistance, the energetic substrates are safeguarded for brain metabolism. Conclusion/interpretation. In the presence of an insulin resistance genotype and westernisation (high carbohydrate diet, increased saturated fat, low fibre and sedentary habit), a high cytokine responder genotype would be prone to deterioration of insulin resistance and, finally, to Type II (non-insulin-dependent) diabetes mellitus and atherosclerosis. For our ancestors, the advantages of a high cytokine responder (eradication of injury) or moderate insulin resistance (protection against starvation) overcame the possible inconveniences of atherosclerosis. We propose that the latter are good adaptations to the environment or “maladaptations” of actual lifestyle to our genome. [Diabetologia (1999) 42: 1367–1374]