A new method for evaluating fruit firmness is presented by an extensive study on Red Delicious apples. An independent measure for assessing the sensitivity of non-destructive methods for measuring fruit firmness is proposed, by taking two time-spaced firmness readings on a fruit sample, at the same locations on each fruit. The non-destructive method for measuring apple firmness in this study is based on vibrational excitation. The conventional destructive method measures the force required to pierce the fruit by a penetrometer tool. A statistical model was developed for objective comparison of non-destructive fruit firmness measurement methods and the conventional destructive technique by the Magness-Taylor tool. It is shown that linear regression and the correlation coefficient depend strongly on the firmness range in the inspected sample. As the firmness range may vary from sample to sample, regression analysis and the correlation coefficient is not suitable for comparisons of firmness measurement methods. The new approach is to quantify directly the distribution of the differences between the readings of the two methods, by a special normalization, which puts the readings of both methods onto the same non-dimensional scale.