Abstract: In view of the difficulty in the detection and evaluation of deep and small holes with a diameter of less than 1 mm and a depth-to-diameter ratio of more than 10∶1, a method of using ultrasonic equipment to detect deep and small holes with a large depth-to-diameter ratiow was proposed. The hole with a diameter of 1 mm and a depth of 100 mm was machined on the experimental workpiece by using the electric spark piercing machine.With the help of the built ultrasonic testing platform, the deep hole was regarded as the internal defect of the workpiece, and the machined deep hole workpiece was detected on the surface of the workpiece. The acoustic path time data obtained by ultrasonic equipment detection was transformed into the coordinate data of deep holes by coordinate processing. Four cylindricity error evaluation models such as maximum inscribed cylinder method, minimum circumscribed cylinder method, least square method and minimum area method optimized by genetic algorithm were used in MATLAB software to analyze the error of the coordinate data of deep holes.The results show that the proposed detection method can successfully detect small holes with large depth-to-diameter ratio. The four cylindricity error evaluation models optimized by genetic algorithm can solve the corresponding cylindricity error, and meet the criterion of minimum error difference of minimum area method.