Abstract: Since the hop band of slow-hopping frequency hopping (FH) signals is large, the sample rate is very high, and the amount of data needing to be processed is large, a time difference of arrival (TDOA) estimation method based on part data split jointed is proposed for the slow-hopping signals, which can get accurate results quickly. There are periodical peaks in the correlation results of FH signals, and it will bring vague estimation results. To restrain the influence of periodical peaks, two methods are proposed. The first is to set the search range of TDOA in a period with the use of pre-information. The second is to make use of the estimation result of single hop signal, and the TDOA search range will be set in a period include the estimation result of single hop signal. The results of computer simulations demonstrate that the proposed methods can restrain the influence of periodical peaks effectively, and get very accurate estimation results when the signal-to-noise ration is higher than 8 dB.