OBJECTIVE Neoadjuvant chemotherapy (NAC) is applied in locally advanced breast cancers (LABCs). Pathological complete response (PCR) after NAC is associated with prognosis. This prospective study aimed to com- pare the predictive value of semi-quantitative parameters obtained by dynamic contrast-enhanced (DCE) magnetic resonance imaging (MRI) and dual-phase 18F-FDG PET/CT in LABC patients receiving NAC. METHODS Thirty-nine patients with LABC underwent DCE-MRI and 18F-FDG PET/CT at baseline, and 38 after 2-3 cycles of NAC (interim). Tumor diameter, spherical volume (SV), angiographic volume, peak signal intensity (PSI), the rapid and medium component of initial rise, and percentage of Type I, Type II, and Type III curves were calculated. SUVmax, total lesion glycolysis (TLG), and metabolic tumor volume (MTV) were measured using adaptive (adp) and 42% thresholding methods in whole-body and late prone images. Baseline and interim studies calculated percentage changes and compared the surgery results, PCR, and non-PCR. ROC curves were obtained to calculate the area under the curve for PCR prediction. Optimal threshold values to discriminate between PCR and non-PCR were calculated. RESULTS Late prone images had higher sensitivity and specificity to detect the residual tumor (91%, 71.4%) than MRI (84%, 37.5%). 18F-FDG PET/CT parameters differed significantly between PCR and non-PCR groups, except for MTV-42 values. Optimal cutoff values were-65% for SV%, 73% for MTV-adp%, and 88% for TLG-adp%. CONCLUSION Semi-quantitative parameters for 18F-FDG PET/CT and volumetric changes obtained with DCE-MRI can predict response to NAC. Percentage changes in SV, MTV, and TLG can identify non-responding patients better than other parameters.