The invention discloses a method for rapidly detecting a contrast sensitivity threshold value, and belongs to the technical field of contrast sensitivity threshold value detection. The method comprises the following steps: S1, selecting a measurement function and a maximum iteration number, and initializing a lookup table of a current iteration number, a prior probability density and a conditional probability; s2, calculating the probability after feedback is obtained next time; s3, predicting the probability of each measurement function; s4, predicting the entropy of the probability density function of the measurement function in the whole parameter space; s5, predicting the expected entropy of each stimulus; s6, finding out the stimulation intensity when the expected entropy is minimum; s7, designing a disturbance operator and then adding the disturbance operator to the stimulation intensity of S6; s8, adding a disturbance operator, and calculating to obtain posterior distribution; s9, calculating new parameter prediction; and S10, if the current iteration number is equal to the input maximum iteration number, outputting a threshold value, otherwise, skipping to the step S2. According to the invention, stimulation can be adaptively optimized and adjusted, the speed of threshold estimation is accelerated, the accuracy is ensured, and the requirement of equipment is reduced.