Abstract: Faced with an increasingly complex electromagnetic environment of space, Synthetic Aperture Radar (SAR) is vulnerable to Radio Frequency Interference (RFI) in the same frequency band, which seriously affects the interpretation of SAR images. The traditional instantaneous spectral notch method is simple to implement, but it conducts interference notch on the time-frequency diagram by threshold value. Therefore, the interference suppression effect is not good because of no considering the residual interference component in the time-frequency diagram. To this problem, this paper presents a method of SAR image radio frequency interference suppression based on time-frequency location and edge notch. First, RFI detection is carried out in the two-dimensional time domain. The time-frequency diagrams are obtained by discrete Short-time Fourier transform(STFT) for those disturbed range echoes. Second, interference points in the time-frequency diagram are extracted by Pauta criterion which are sorted roughly by Depth First Search algorithm. Then different inference strips are obtained by detection and unlocking of decussation. After detecting the edges of strips, the edge notch of interference strips is achieved. Last, the suppressed time-frequency diagram is inversely converted to range echo. The finally suppressed SAR image is obtained by imaging echoes which are composed of the suppressed range echoes and the originally undisturbed range echoes. The experimental results on simulated and measured SAR data show that the edge notch method is suitable for different scenarios and various types of RFI. The residual interference components are effectively eliminated under different jam-to-signal ratios which certificates that the method presented in this paper achieves better suppression efficiency than frequency notch method and instantaneous spectral notch method.