The smoothing method called extreme envelope average (EEA) was recently proposed to smooth one-dimensional signals. Its basic principle is based on the central tendency of successive averages between extreme envelopes, maximum and minimum, after a specific number of steps of an iterative process. Energy levels are lost after each iteration and the most important signal structures (low-frequency components) are preserved. There are no restrictions on the linearity and harmonic content, making it a useful tool for signals which present abrupt changes (such as sharp edges) along their curvatures. Its efficiency is highly related to the identification process of local maximums/ minimums and the interpolation process. In this work, the more robust version of EEA called robust extreme envelope average REEA is used to reduce noise of beam mode shapes from numerical modal analysis. The used modes were numerically corrupted by White Gaussian noise before applying the smoothing technique.