The invention discloses a method for establishing a cerebral apoplexy recurrence prediction model. The method comprises the steps of 1, establishing a comparison database; 2, establishing a cerebral apoplexy recurrence prediction system; and 3, integrating the comparison database established in the step 1 into an intelligent analysis module. The invention further discloses a cerebral apoplexy recurrence prediction device which comprises a computer, a heart rate tester, a blood pressure measuring instrument, an oximeter, a thermometer and a blood lipid instrument; the computer comprises a host,a display, a keyboard and a mouse; a cerebral apoplexy recurrence prediction model is installed in the host, and the display, the keyboard and the mouse are electrically connected with the host through data lines. The method and the device are proposed for elderly patients of 75 years old or above, are suitable for hospitals and families, can avoid troubles caused by the fact that the elderly patients need to frequently perform cerebral apoplexy recurrence prediction, solve life troubles, effectively guarantee the life rhythm of the elderly patients, and further improve the life quality of the elderly patients.