Objective　Chinese medical specialists asserted that the drumstick spinal process(DSP) originated from the bony union of the normal L5 spinal process and isolated S1 spinal process from sacral spina bifida occulta.  The imaging diagnosis is generally established based on the anteroposterior conventional radiograph image. This study was designed to verify the assertion by analysis of DSP images and direct visual inspection.　Methods　Cases met DSP imaging diagnosis criteria of plain radiographs from December, 2004 to July, 2016 were enrolled in the study, anteroposterior and lateral conventional radiographs of relative cases were observed to find DSP’s morphologic characteristics. If CT scanning or MR scanning was performed, sagittal reconstructed images at midline were observed to find the image evidence of DSP’s components. If open operation was performed, direct observation of DSP with naked eye was carried out.　Results　16 cases satisfied the DSP’s imaging diagnosis on plain radiographs which showed the distal end of the spinal process connected with L5 laminae reached the central defective area of sacral spina bifida occulta were enrolled. The contour of DSP at the lateral radiograph was clearly visible in 12 cases, which exhibited fin morphology, among whom, 7 had smooth rim while 4 had a dent at the inferoposterior edge.  CT scanning was done at three cases, two had sagittal CT images reconstructed which detected a narrowed region near the lower portion of DSP corresponding to the dent location found on plain radiographs and sclerotic bands could be detected adjacent to the narrow region distally and proximately. MR scanning was done on three cases. In two cases, sagittal MR images at the midline exhibited low T1WI and T2WI signals at the narrow area of DSP corresponding to the relative area at CT images and high T1WI and T2WI signals at bilateral margins adjacent to the narrow area distally and proximately, while in one case, the signals of the narrow area of DSP were similar to that of the normal L5S1 interspinal ligament, thus making it difficult to discern the existence of DSP. Open operation was performed on 2 cases, a dent could be found at inferoposterior edge of DSP in one case by naked eye. 　Conclusion　This study can support the assertion that DSP comes from the bony union of the normal L5 spinal process and the isolated S1 spinal process of sacral spina bifida occulta. The narrow area at lower portion of DSP detected by reconstructed sagittal CT images is the union location.