A new method of detecting short track defects from in-service vehicles is developed using successive impact of the leading and rear wheels on short track defects. The initial moment and the end moment as a rotation of wheel, which are determined through integration of the operational velocity of train, are used to truncate and divide the time history of leading and rear axle vertical accelerations. The truncated acceleration signals between initial moment and the end moment were transformed by frequency sliced wavelet transform (FSWT) to get the time-frequency characteristic of wheel-rail impact. The time axis was replaced by displacement axis signifying wheel circumference in the time-frequency figures to achieve time-frequency feature locations as a rotation of wheel. The model of detecting short track defects was constructed by comparing the location of wavelet energy peak in continuous three circles. The effectiveness of way has been proved through much dynamics simulation data, with high real-time, high accuracy and a good engineering applicability.