A long-term study of two diets on the progression of diabetic retinopathy was performed on 102 patients. Diet 1 was composed of carbohydrates 50 cal%, saturated fats 35 cal%, proteins 15 cal%; diet 2, of carbohydrates 45 cal%, fats 40 cal% (are third linoleic acid) and proteins 15 cal%. The linoleic acid content of diet 2 was 4 times that of diet 1. The diabetics were matched 2 by 2 according to 10 criteria and set at random on one of the two diets. During the observation period of 5 years, 3 patients of diet 1 died of a heart infarction. Of the remaining 96 patients (51 males and 44 females); of diet group 1 (linoleic acid poor), 62% of the males and 55% of the females showed retinopathy. Of diet group 2 (linoleic acid rich), 27% of the males and 32% of the females showed retinopathy (p < 0.00l). A strong inverse correlation was found between progression of retinopathy and serum cholesteryl linoleate. Of diet group 1, 11 males and 7 females showed signs of cardiac ischaemia, of diet group 2, 3 males and 2 females (p < 0.025) did so. Large differences were seen for the GTTs and corresponding insulin levels between males and females. While the females of diet group 2 improved remarkably in respect of GTTs and reactive serum insulins, a slow deterioration of both parameters was seen for diet group 1. Regarding the males, no differences between the two diet groups were seen. The improvements due to the treatment were similar. The serum lipid responses were significant for the females, diet group 2 showed the lowest levels; however, no differences were seen for the males. In most cases of both sexes, platelet aggregation was increased, no effects of linoleic acid was seen: an explanation for the contradictory facts was not given.