The invention provides a fuzzy reasoning and taylor expansion-based online load prediction method of a forging press. The method comprises the following steps of (1) initializing parameters of a taylor expansion model and a membership function; (2) building various orders of continuous conductivity indexes of taylor expansion and calculating the membership of various orders of taylor expansion models by employing the various orders of continuous conductivity indexes as the parameters of the membership function of fuzzy reasoning; (3) carrying out parameter identification of various orders of taylor expansion models according to the online perception data; (4) normalizing the memberships of various orders of taylor expansion models and determining a prediction result according to the normalized memberships of various orders of taylor expansion models and the calculation result; and (5) predicting the load of the press at the next moment through online perception of the load of the forging press and the displacement of an upper beam. According to the method, the load of the forging press can be accurately and quickly predicted online, and a technical support is provided for accurate control on the load of the forging press.