The aim of this study was to assess whether positron emission tomography (PET) or magnetic resonance imaging (MRI) could obviate the need for surgical staging in patients with locally advanced cervical carcinoma being planned for radiotherapy (RT). Imaging findings were compared to surgical staging in 27 patients including three with recent resection of the primary tumor. Both PET and MRI visualized all 24 residual cervical tumors. Primary tumor volume, as measured by MRI scan, ranged from 1.25 cc to 140 cc. In 24 patients evaluable for pelvic nodal status, PET had sensitivity, specificity, and positive and negative predictive values of 83%, 92%, 91% and 85%, respectively, with 88% accuracy. MRI detected only six in 12 (50%) patients with confirmed pelvic nodal disease, all of which were also seen by CT and PET, with an overall accuracy of 75%. PET detected only four in seven (57%) cases with confirmed para-aortic (PA) involvement. All histologically confirmed sites not visualized on PET were <1 cm. Without surgical staging, six in 10 (60%) patients with histologically proven pelvic nodal disease would not have received pelvic boost if guided by MRI alone, compared to two in 10 (20%) patients guided by PET alone or in combination with MRI. All four patients with positive PA on PET were confirmed on histology or clinical follow-up, including one case that proved to be a false negative one on surgery. However, in three cases, PET would have yielded an inadequate radiation volume. In conclusion, the positive predictive value of PET in the pelvis and para-aortic region appears sufficient to obviate lymph nodal sampling, but sampling is still required to exclude small-volume disease cranial to sites of abnormality on PET. MRI has insufficient accuracy for nodal staging to impact management.