Seismic processing characterizing thickness and borders of thin inter-beds has gradually evolved from post-stack migration to pre-stack migration, and the latter considers both vertical and lateral resolutions. As the key processing methods for improving vertical and lateral resolution, conventional deconvolution and pre-stack time migration (PSTM) are not simply dominated by the estimation and compression of the wavelet because of its instability. Therefore, considering the variations of wavelet frequency before, during and after PSTM can obtain good common reflection point (CRP) gathers and imaging profiles of thin inter-beds. Based on the frequency characteristics of the wavelet before, during and after PSTM, a joint high-resolution processing method for thin inter-beds is proposed in this paper, including inverse Q filtering for high-frequency compensation before PSTM, optimum weighting Kirchhoff PSTM for preserving high-frequencies during PSTM, and wavelet harmonizer deconvolution for consistent processing and frequency-band broadening after PSTM. An application to real data characterized by mudstone beds in the Oriente Basin proved that the joint high-resolution processing method is effective for determining the thickness and borders of thin inter-beds and is favorable for subsequent reservoir prediction and seismic inversions.