It is uncertain whether increased arterial pressure alone or increased arterial pressure combined with some other factor is responsible for the development of malignant-phase hypertension. Our object was to test this comparing two groups of rats with hypertension of different mechanism but of similar duration and degree. Two experiments were done in hypertensive rats. In the first blood pressure was measured in the tail, in the second by intra-arterial catheter and pressure transducer. In the first, hypertension was produced in 30 male Sprague-Dawley rats by unilateral nephrectomy followed by thrice-weekly injections of deoxycorticosterone (12.5 mg) and substitution of 1% NaCl and 0.2% KCl for their drinking water. After four weeks 28 rats survived and systolic blood pressure had risen to 220 mmHg. The survivors were paired by blood pressure and randomly allocated either to continued DOC and salt or to a regime in which DOC was stopped and tap-water was substituted for the NaCl-KCl solution (post-DOC rats). Blood pressure remained similar in the two groups thereafter, but in every case it was the DOC-salt animal of the pair which died first. Fibrinoid arteriolar lesions of malignant-phase hypertension were significantly commoner in DOC-salt animals and, before death, they had more pronounced features of microangiopathic haemolytic anaemia. The second experiment was the same in design as the first except that two weeks after randomization the similarity of blood pressure was confirmed by measurement of intra-arterial pressure in eight pairs of DOC-salt and post-DOC rats. We conclude that increased arterial pressure is not the only factor influencing development of malignant-phase hypertension.