Male stroke-prone spontaneously hypertensive rats (SHRSP) were fed 4% NaCl diets containing either 0.75% normal K or 2.11% high K, starting at 6 weeks of age. After 8 months on these diets 69% of 58 SHRSP rats on 0.75% K had died, whereas 2% of 95 rats of 2.11% K died, a 98% reduction in mortality, P less than 0.000 001. After 20 weeks the daytime and night-time blood pressure (BP) of each rat were measured intra-arterially. We selected two groups precisely matched for BP. One matched SHRSP group (BP 182 mmHg) ate the 0.75% K diet and 30 of 47 rats died (64% mortality). The other matched SHRSP group (BP 182 mmHg) ate the 2.11% K diet, and two of 35 died (6% mortality, a 91% reduction of mortality, P less than 0.0001). Seemingly, the striking reduction in mortality rate with the 2.11% hig-K diet does not depend on a lowering of BP. High-K diets do not change body Na or K. The dry weight of mesenteric arterioles was reduced by 22% on 2.11% K diet versus 75% K (7.5 versus 9.7 mg) (P less than 0.001), indicating a greatly reduced hypertensive hypertrophy. In nine surviving SHRSP on 0.75% K, 13 of 36 brain hemisphere slides (four slides per rat) showed infarcts (36%). In 11 surviving SHRSP on 2.11% K, one of 44 brain slides showed infarcts (2%, a 94.5% reduction, P less than 0.0001). Brain haemorrhage was reduced by 92% on the 2.11% K diet. High-K diets allow cerebral arteries to carry very high BPs without sustaining damage to the artery wall, thereby drastically reducing brain infarcts and lowering the death rate.