This work presents the results of the behavior of the inductive proximity sensor at the measurement of oscillation in hydroelectric generators from the generation system of Furnas Centrais Elétricas S.A.. To develop this study, two (2) inductive proximity sensors were used. The methodology consisted of calibrating statically and dynamically the inductive proximity sensor. For this, it was used a test mockup: a system composed by an engine with variable speed and a disc with variable and controlled eccentricity connected to the axle. The eccentricity of the test mock-up was adjusted and thirteen (13) points were marked on the disc, every 30°. The static calibration of the test mock-up has been made, by measuring the displacement in each one of these thirteen (13) points with a calibrated dial indicator. For the static calibration of the sensor, the sensor itself was positioned in front of the test mock-up disc at a known distance, the axle was dislocated manually for each one of the points and the output signal of the sensor was measured with a calibrated signal analyzer. The curve of the static calibration then was determined with the respective uncertainty of the measurement. For the dynamic calibration, the same system was run in the nominal speeds of the hydroelectric generators from Furnas Centrais Elétricas S.A. (90 rpm, 120 rpm, 150 rpm and 180 rpm). The response of the sensor was then compared at the different rotations, being determined the attenuation of the signal due to the increase of the operation frequency. The static and dynamic data were evaluated, and the uncertainties determined. The static calibrations carried through by the procedure of FURNAS and by this work produced the same results, considering the range of measurement uncertainty. To reduce the calibration uncertainty of the FURNAS procedure, it is proposed that its results be adjusted by a polynomial of second degree. The dynamic calibration carried at several frequencies showed that there is no difference between the static and dynamic values, even though the dynamic calibration has a more elevated uncertainty. To simplify the calibration of the sensors, it is proposed then that it is carried statically only.