Background Clinicians are unsure if radiography is needed to confirm correct positioning of feeding tubes inserted with assistance from an electromagnetic system. Objectives To compare radiographic reports of feeding tube placement with images generated by an electromagnetic feeding tube placement device. Methods The medical records of 200 consecutive patients who had feeding tubes inserted with assistance from an electromagnetic feeding tube placement device were reviewed retrospectively. Radiographic reports of tube site were compared with images generated by the device. Results Radiographic evidence of tube sites was available in 188 cases: 184 tubes were located in portions of the gastrointestinal tract. Ninety of the 188 tubes were situated in the optimal site (distal duodenum or jejunum) radiographically. Images generated by the electromagnetic device were available in 176 cases; of these, 52 tubes appeared to end in the expected left lower quadrant. Tubes shown on radiographs to be in other sites also occasionally appeared to end in the left lower quadrant. Nurses using the device did not recognize 4 of the 188 tubes (2.1%) that were inadvertently placed in the lung. No consistent pattern of quadrant distribution was found for tubes positioned in the stomach or proximal duodenum. Conclusions Images generated by the electromagnetic tube placement device provided inconsistent results regarding tube location. A small percentage of seriously malpositioned tubes were not detected by using the electromagnetic device. These findings do not support eliminating radiographs to confirm correct tube placement following use of an electromagnetic tube placement device.