Purpose: A measurement device for noninvasive and simultaneous control of antennas during regional radiofrequency (rf) hyperthermia and, subsequently, the estimation of the power distribution in the interior of patients are essential preconditions for further technological progress. Aiming at this, the feasibility of an electro-optical electric field sensor was investigated during clinical rf hyperthermia.Material and Methods: The electro-optical electric field (E-field) sensor is based on lithiumniobate crystals and the Mach–Zehnder interferometer structure, and was tested in an earlier phantom study. For this study, a mechanical scanning device was developed allowing the registration of the E-field during clinical application. Data were recorded along a curve in the water bolus of the SIGMA 60 applicator of the annular phased-array system BSD-2000 (BSD Medical Corp., Salt Lake City, UT) close to the base points of the flat biconical dipole antennas. The results were compared with modeling calculations using the finite-difference time-domain (FDTD) method. For the latter, different antenna models were assumed. For systematic registration of the E-field curves in amplitude and phase, we employed an elliptical lamp phantom with fat-equivalent ring (filled with saline solution) and an elliptical polyacrylamide phantom with acrylic glass wall. Further measurements were carried out during the treatment of 5 patients with 20 hyperthermia treatments.Results: Data of both phantom and patient measurements can be satisfactorily described by the FDTD method, if the antenna model is refined by taking into account the conical form of the dipoles and the special dielectric environment of the feeding point. Phase deviations can be entered ex posteriori for correction in the calculation algorithm. A comparison of amplifier power measurement (forward and backward power) and bolus E-field scans near the antenna base points demonstrates that E-field measurements between antennas and patient are a necessity for the appropriate characterization of antenna radiation properties. These measurements are sensitive to variations of the lossy medium in position and shape, and can be correctly predicted with current models. However, the differences between different patients are moderate and unspecific in both calculations and measurements, with fluctuations at maximum of 30° in phases and 40% in amplitudes.Conclusions: The measurement method presented here turned out to be a practical tool for online registration of E-fields in phases and amplitudes along arbitrary curves in a water bolus or phantom. It can be utilized to evaluate antenna design and modeling calculations and leads, thus, to a better understanding of complicated multiantenna systems. In clinical routine, it can be employed as input for patient-specific hyperthermia planning and, finally, for the realization of online control with subsequent optimization of the power distribution in the patient.