Abstract: Mercury injection capillary pressure (MICP), rate-controlled porosimetry (RCP) and nuclear magnetic resonance (NMR) have limitations in describing the characteristics of microscopic pore structure characteristics of extra-low permeability sandstone reservoirs, and the results are not completely consistent with the observation of thin sections and scanning electron microscopy. Five ultra-low permeability sandstone reservoir samples were collected from the Heshui area of Ordos Basin. A collaborative multi-method for characterizing pore throat structure was proposed in order to describe the detailed characteristics of pore size distribution. MICP and NMR combined with MICP were used to obtain pore connectivity. The connectivity ratio of adsorption throat, micro throat, fine throat, middle throat was calculated. Nuclear magnetic resonance data were used to achieve the conversion of transverse relaxation time to pore throat radius. The specific surface area was calculated by MICP, and the relaxation rate was calibrated by RCP and the T₂ spectrum. The synergistic calculated pore throat distribution results were multiplied by the corresponding pore throat connectivity ratio to obtain the spatial distribution curves of throat and pore connectivity at different scales. The results showed that the adsorption throat connectivity ratio was the lowest, and the ratio of other sizes was relatively higher, but the difference is not significant. The throat radius ranged from 0.003 to 3.661 μm, which was greater than constant rate-controlled porosimetry test results. The pore radius ranged from 0.8 to 91.4 μm and the pore-throat ratio ranged from 16.4 to 58.6 μm, both of which were smaller than rate-controlled porosimetry test results. The final calculation results were basically in accordance with the observed results of cast thin section and scanning electron microscopy. It showed that the collaborative calculation of multiple methods overcomes the superposition of throat and pores on high-pressure mercury-injection and the calculation error of rate-controlled porosimetry, which is closer to the true state of reservoir.