The invention discloses a non-linear Volterra filtering optimization method based on a contribution factor, and the method comprises the steps: 1, obtaining an original output signal of a non-linear Volterra filter; 2, setting the weighting coefficient of each node of the non-linear Volterra filter as zero, forming sub-filters, and obtaining an output signal of each sub-filter; 3, calculating the contribution factor of each node according to the output signal and the original output signal of each sub-filter; 4, comparing the contribution factor of each node with a set threshold value delta, and determining the optimal weighting coefficient of the non-linear Volterra filter; 5, forming an optimized filter according to the optimal weighting coefficient. The method aims at a problem that a conventional non-linear Volterra filtering method is large in calculation burden, proposes a method for searching and deleting the redundant nodes according to the attribution of each node to the corresponding output signal, deletes the redundant nodes with the minimum contribution to the output result under the condition that the filtering performance is not affected, can further reduce the calculation amount, and reduces the occupied resource of engineering implementation.