This study aims to detect micro defects in multilayer ceramic capacitors (MLCC) using a new signal processing method for electromechanical (EM) responses. Microscopic defects within MLCC structures, not screened during the manufacturing production stage, can cause damage to electronic products and further lead to overall system failure. Therefore, it is very important to secure reliable defect detection technology. However, conventional methods, such as X-ray and ultrasound tests, are destructive, cost-ineffective, and inaccurate. In this study, we applied a new signal processing method based on the Hilbert-Huang transform (HHT) of the EM response for the nondestructive screening of MLCCs with hidden defects. We first observed that infinitesimal, irregular oscillation in the instantaneous frequency in the transient response when the MLCC had defects. Therefore, we introduced a new signal processing method that takes the Fourier transform (FT) of the weightedinstantaneous frequency (WIF) concerning the frequency-oscillatory rate, to capture the infinitesimal change in the HHT signal. Specifically, we confirmed that subharmonic terms, possibly caused by the nonlinear effect, arose in the WIF when the target MLCC had a micro defect. The method was first validated through signals obtained from transient finite element analysis, followed by a comparison with the test data from actual MLCCs. The increased subharmonic terms in the WIF could be an effective indication of the defect in the MLCC. This study will provide a new methodology for detecting microdefects in MLCC.