Improving the characterization of deep carbonate reservoirs requires developing a clear understanding of nature and distribution of their constituent microfacies and associated pore types. These aspects have been little studied in the middle-lower Ordovician of the Tahe Oilfield in the Tarim Basin in large part due to the limited available cores and relatively poor seismic data. Formation MicroImage (FMI) logging provides a bridge to connect the core data and seismic data to facilitate study of the distribution of microfacies and pore types. Based on analysis of the FMI logs from eight wells (each calibrated by comparing with core samples, thin sections, and conventional well logs), five FMI-based microfacies have been established: (1) intershoal sea microfacies (ISMF), (2) low-energy shoal microfacies (LSMF), (3) high-energy shoal microfacies (HSMF), (4) lagoon microfacies (LMF), and (5) tidal flat microfacies (TFMF). Using the FMI-based identification of microfacies, it has been proposed that the Yingshan Formation was deposited in restricted-platform (characterized by LSMF, HSMF, LMF, and TFMF) and open-platform (characterized by ISMF, LSMF, and HSMF) environments. Three pore types, with the potential for reservoir quality porosity, have been identified in FMI logs: pores (laminated and isolated pores), vugs, and fractures (dipping shear and conjugate fractures). Statistical analyses found that, in comparison with other microfacies, HSMF is more favorable for the development of the above three reservoir pore types, vugs being the most abundant. This study shows the effectiveness of FMI images in the identification of carbonate microfacies and reservoir pore types, and in the building of high-resolution 3D geologic models that identify high-quality reservoir zones.