The invention discloses a flash memory service life prediction method based on support vector regression. The method comprises the following steps: (1) obtaining the characteristic quantity of a flashmemory to be tested, wherein the characteristic quantity comprises the programming time, the reading time, the erasing time, the current, the power consumption, the threshold voltage distribution, the storage block number, the storage page number, the programming / erasing period number currently experienced by the flash memory, the condition error page number, the condition error block number, theerror bit number and the error rate; (2) processing the characteristic quantity to obtain an operation result; And (3) forming a set by the characteristic quantity and the operation result, and performing support vector regression operation by taking the subset in the set as the input of a support vector regression model to obtain a service life prediction value of the flash memory correspondingto the characteristic quantity. The residual service life of the flash memory can be predicted, so that a flash memory storage device user knows the loss state of the memory in the using process of the device, and data loss caused by failure of the memory unit is avoided.