To the Editor:—We read with interest the correspondence by Zimmer et al. Although we agree with their conclusion that human error related to magnetic resonance imaging (MRI) use can only be minimized by adequate training, we believe lessening the risks of MRI technology itself is of equal importance. We feel it is time to stress the importance of “anesthesia-compatible” MRI, rather than putting all the emphasis on anesthesiologists adapting to the needs of the MRI machine. Anesthesiologists and patients are now forced into working under conditions that are far less than optimal in MRI suites that are cold and dark, have noisy equipment and facilities, and are often located far away from the main operating area. There are three aspects of MRI that are important to the anesthesiologist: 1) avoidance of materials and equipment that will be attracted to the MRI machine, 2) avoidance of anesthetic devices that interfere with the function of the MRI machine, and 3) avoidance of MRI interference with the patient and the functioning of materials and devices used for anesthesia. This third aspect is often neglected: most MRI machines are not “anesthesia compatible.” The first point is obvious, well known, and scary but can be handled with a little experience. It is, however, extremely important for everyone to realize that complete elimination of the use of ferromagnetic materials in devices used in MRI suites is not feasible and is sometimes impossible, as was made clear in the letter by Zimmer et al., among others. We have successfully dealt with this problem by anchoring all devices that have ferrous materials in them to a movable ceiling pendant system with a predetermined limited range of movement. Installation of metal detectors (similar to those used in airports) at the entrance to MRI suites can help to some degree. The second point involves the use of equipment such as ventilators and infusion pumps for treatment and various patient monitors, personal computer-related devices, and local area network connections for medical information. MRI technology now frequently forces the anesthesiologist to discontinue the use of these devices during MRI activity despite their importance for patient safety. Alternative “MRIcompatible” devices are not always available, functional, or suited for critically ill patients, causing anesthesiologists to make an uncomfortable choice between the continuity of safety of treatment and MRI diagnosis. The third aspect is the most difficult to resolve. We think more attention should be focused on alleviating this problem although some MRI properties, such as magnetic attraction, electric shock, or heating as a result of radiofrequency pulsing, seem to be inevitable. Anesthesiologists have been forced to adapt to MRI technology, raising a never-ending list of incompatibility issues. While working to build a new MRI suite, we realized that although there were few technical difficulties to overcome, lack of awareness of the issues involved with traditional MRI was playing a key role in holding back the development of more patient-friendly MRI technology. Companies we attempted to work with that already make both MRI and anesthesia-related equipment did not seem to find safety for patients under anesthesia during MRI a compelling enough reason to consider revising their MRI devices. It would be much more cost effective and safe to improve MRI machines and their installation, including the architectural design of MRI suites, than it would be to carry out patchwork renovation of numerous patient care devices. The demand for anesthesia care in MRI suites continues to increase as interventional procedures using MRI continue to increase in frequency. Time spent in MRI suites will only become longer. Anesthesiologists, as advocates for patients, should actively voice their concern to improve MRI technology not only in terms of radiologic diagnostic function but also in terms of working environment, duration of examination, and, most importantly, patient safety. We must seek solutions for safer anesthesia delivery. We should stop being cursed by the need for “MRI-compatibility” and start actively implementing an “anesthesia-compatible” MRI environment. To achieve this goal, anesthesiologists should be involved from the beginning of the conceptual design of MRI suites.