Awareness of the ancestral human diet might advance traditional nutrition science. The human genome has hardly changed since the emergence of behaviourally-modern humans in East Africa 100-50 x 10(3) years ago; genetically, man remains adapted for the foods consumed then. The best available estimates suggest that those ancestors obtained about 35% of their dietary energy from fats, 35% from carbohydrates and 30% from protein. Saturated fats contributed approximately 7.5% total energy and harmful trans-fatty acids contributed negligible amounts. Polyunsaturated fat intake was high, with n-6:n-3 approaching 2:1 (v. 10:1 today). Cholesterol consumption was substantial, perhaps 480 mg/d. Carbohydrate came from uncultivated fruits and vegetables, approximately 50% energy intake as compared with the present level of 16% energy intake for Americans. High fruit and vegetable intake and minimal grain and dairy consumption made ancestral diets base-yielding, unlike today's acid-producing pattern. Honey comprised 2-3% energy intake as compared with the 15% added sugars contribute currently. Fibre consumption was high, perhaps 100 g/d, but phytate content was minimal. Vitamin, mineral and (probably) phytochemical intake was typically 1.5 to eight times that of today except for that of Na, generally <1000 mg/d, i.e. much less than that of K. The field of nutrition science suffers from the absence of a unifying hypothesis on which to build a dietary strategy for prevention; there is no Kuhnian paradigm, which some researchers believe to be a prerequisite for progress in any scientific discipline. An understanding of human evolutionary experience and its relevance to contemporary nutritional requirements may address this critical deficiency.