The invention discloses a strong signal screening method for SRS combined drug adverse reaction signals, comprising: acquiring data through the spontaneous drug adverse event reporting system SRS, and obtaining combined drug adverse reaction signal set X={D 1 ,D 2 ,...,D x},D x is the drug combination judged as an adverse reaction signal; assuming that the signal exclusion set in the t-th cycle is Y t , then select the signal D with the largest current signal metric st As a strong signal, and put the signal into the signal exclusion set Y t Form a new signal exclusion set Y t+1 , and then start the t+1th round of circulation, and so on, to get the set of signals with the largest signal metric in each round of circulation, that is, the strong signal set S={D s1 ,D s2 ,...,D st}, screen out strong signals with significant case data basis from a large number of combined drug adverse reaction signals, which improves the quality of signal recommendation.