At high sound pressure levels, actual body vibrations (noise-induced vibrations) are induced by low-frequency noise. The purpose of this trial study was to show that considering the effects of noise-induced vibration is effective in evaluating high-level low-frequency noise. Using the A-weighted sound pressure level and the W_k-weighted vibration acceleration level of noise-induced vibration measured on the chest as independent variables, empirical evaluation indices (HLLF1, HLLF2 and HLLF3) for evaluating the unpleasantness caused by high-level low-frequency noise were estimated. The HLLF indices were found to be able to evaluate the unpleasantness caused by high-level low-frequency noise better than the A-weighted pressure level. In addition, the slopes of tentative frequency-weighting characteristics corresponding to the HLLF indices were estimated to be gentler than that of the A-weighting characteristic within 25-50 Hz, which was consistent with many previous results that indicated that noise content at lower frequencies should be given more importance when evaluating low-frequency noise Although there are several areas where the HLLF index needs to be improved before it is put in practical use, the results of this study suggest that high-level low-frequency noise could be more effectively evaluated by taking into account the effect of human body vibration.