The invention relates to a distribution network short-circuit calculation method based on matrix transformation, and belongs to the field of power system distribution network short-circuit calculation. The method comprises the steps of 1) establishing a distribution network three-phase system model based on the characteristic of the distribution network three-phase asymmetry; 2) forming a distribution network node admittance matrix by use of the distribution network three-phase system model; 3) performing load flow calculation for a distribution network by use of a three-phase Newton-Raphson method; 4) enabling a centralized load to be equivalent to constant impedance and merging into a distribution network model, and modifying the node admittance matrix; 5) establishing a node voltage equation of the distribution network; 6) selecting a short-circuit node and a short-circuit type and whether a neutral point in a balance node is grounded determine the corresponding boundary condition; and 7) performing matrix transformation for the network node voltage equation and solving the equation to determine whether a relay protection device in the distribution network operates or not. The method takes the load as constant impedance load with constant impedance, merges the equivalent impedance of the load into the node admittance matrix, and greatly reduces the number of variables in the equation.