Objective To analyze the application value of multi-slice spiral CT image postprocessing techniques in the diagnosis of pulmonary embolism. Methods The clinical data of 48 patients with pulmonary embolism treated in the People's Hospital of Gansu from July 2012 to June 2015 were analyzed retrospectively. All patients were examined with multi-slice CT after admission, and images were processed by the maximum intensity projection (MIP), multi-planar reconstruction (MPR) and volume rendering (VR) postprocessing techniques. The value of different postprocessing techniques in the diagnosis of pulmonary embolism was analyzed. Results There were a total of 491 emboli involving different levels of pulmonary artery branches in 48 patients with pulmonary embolism. The detection rates of pulmonary segment artery and sub-segment artery emboli by VR were lower than those by MPR and MIP, and the overall detection rate of the pulmonary embolism emboli by VR was 81.1%, lower than that of MPR and MIP reconstruction (P<0.05). Eccentric filling defect was the most common direct sign of 491 pulmonary embolism emboli, followed by complete embolism and central filling defect, and mural thrombus was rare. Indirect signs included pulmonary infarction, pulmonary hypertension, pulmonary artery thickening, pleural effusion, limited sparse lung texture and bronchial artery dilatation. Conclusion The sensitivity and specificity of multi-slice spiral CT are high in the diagnosis of pulmonary embolism, safe and non-invasive. The image quality and resolution are high. It can be combined with MPR and MIP postprocessing techniques to improve the detection rate of sub arterial emboli.