Abstract Background The efficacy of neoadjuvant chemoradiotherapy (NACRT) for resectable and borderline resectable pancreatic cancer is important for predicting outcomes after radical surgery, but few clinical indicators predict outcome before resection. This study examined the utility of FDG-PET in predicting the efficacy of NACRT and outcome after radical surgery. Methods Eighty-three pancreatic cancer patients who underwent FDG-PET before and after NACRT and had positive standard uptake values (SUVs) before NACRT were enrolled in this study. Peri-operative clinical factors, including FDG-PET findings, were examined to predict the efficacy of NACRT and outcome after surgery. Results Evans grade I, IIA, IIB, III, and IV was determined in 11, 31, 27, 11, and 3 patients, respectively. The maximum SUVs after NACRT (post SUV-max) and tumor size were significantly decreased compared to pretreatment values (p<0.001 and p=0.007, respectively). The post SUV-max and regression index were significantly related to grade III/IV (p=0.04 and p<0.001, respectively), but only the regression index predicted NACRT efficacy (p=0.002). The AUC of the regression index for the detection of grade III/IV was 0.822, and 13 of 14 grade III/IV patients were picked up using 50% as the threshold (p<0.001). Patients with a regression index >50% had a significantly better prognosis after radical resection than patients with <50% (p=0.032). Regression index as well as pathological lymph node status and resectability status were independent prognostic factors in multivariate analysis (exp2.086, p=0.043). Conclusion The regression index is potentially a good indicator of the efficacy of NACRT and outcome after radical resection for pancreatic cancer.