In the advantages of low memory requirement and high flexibility, sub-aperture method is effective in solving the problem of azimuth spectrum aliasing in sliding spotlight SAR imaging. Image quality and imaging efficiency are affected by sub-aperture partition. However, sub-aperture division usually relies on experience value, which cannot be given with a clear theory analysis at present. Therefore, spectrum gaps may appear in the squint imaging processing if sub-apertures are not properly divided. First, we introduce sliding spotlight SAR model and analyze its Doppler frequency history. A sub-aperture imaging algorithm for squinted sliding spotlight SAR is proposed based on azimuth frequency domain scaling. Then according to two-dimensional spectrum of the adjacent sub-apertures, we investigate the sub-aperture spectrum stitching principle and derive the two equations, thus determine sub-aperture length and overlap rate. Afterward the flow for partitioning sub-apertures is given. Finally, effectiveness of the sub-aperture partition method and squinted imaging algorithm is verified by simulation experiments and real data.