Background Pathologic complete response has been proven to have oncological benefits for locally advanced rectal cancer treated with chemoradiation therapy. The aims of this study are to analyze and determine the factors to predict pathologic complete response for patients treated with preoperative neoadjuvant therapy. Methods Patients with biopsy-proven, locally advanced rectal cancer were treated neoadjuvantly followed by radical surgical resection. Tumors were re-assessed after completing chemoradiation, including pelvic magnetic resonance images, colonoscopic examination, and re-biopsy. The results of examination were compared with the final pathologic status. Results A retrospective chart review of 166 patients was conducted. Twenty-five patients (15.1%) had pathologic complete response after chemoradiation. The 5-year overall survival rates were better in the complete response group than the residual tumor group (91.1% vs. 70.8%; P  = 0.047), and there were also significant differences in the 5-year disease-free survival rates between these two groups (91.1% vs. 70.2%; P  = 0.027). The prediction rates for pathologic complete response by re-biopsy, magnetic resonance images, and colonoscopy were 21.4%, 33.3%, and 53.8%, respectively. In addition, when we further combine the results of colonoscopic findings and re-biopsy, the prediction rate for pathologic complete response reached 77.8% ( P  = 0.009). Conclusions Combining the results of the re-biopsy and post-treatment colonoscopic findings, we can achieve a good prediction rate for pathologic complete response. Post-treatment magnetic resonance images are not useful tools in predicting tumor clearance following chemoradiation.