Split-pit in peach fruit is a problematic disorder. Split-pit fruit cannot be detected based on external appearance, and contamination of fruit by split-pit reduces its reliability in the marketplace. Here, we demonstrate that split-pit fruit can be identified by a nondestructive acoustic vibration method and a unique approach based on the ratio of the third (f 3 ) to the second (f 2 ) resonant frequency. The response-resonant frequency spectra showed that the peaks of f 2 frequencies in split-pit fruit were shifted to much lower values than those in normal fruit, whereas those of f 3 frequencies showed only small shifts. The calculated f 3 /f 2 ratios in most normal fruit were in the range of 1.35-1.4, whereas those in split-pit fruit were 1.45-2.0. Analysis of more than 300 fruit samples revealed that by setting the f 3 /f 2 cut-off value at >1.45, 95% of split-pit fruit in the fruit samples were detected, whereas only 1.5% of normal fruit were missorted as split-pit fruit. A model for simulating the vibration properties of peach fruit was developed by using the finite element method. The simulated vibration patterns showed that f 3 /f 2 values were increased by the insertion of split pit, indicating that, at least partially, the observed high f 3 /f 2 values in split-pit fruit directly reflected split-pit occurrence. These results clearly demonstrate that the use of f 3 /f 2 ratios obtained using an acoustic vibration method can effectively detect fruit with split-pit. The possibility of installing acoustic vibration devices in peach sorting lines and the application of portable devices to unpicked fruit on the tree are discussed.