The potential for increased rainfall, accompanied by strong winds, has the potential to cause floods, landslides, tornadoes to hit various regions in Indonesia, from Sabang to Merauke come and going. Trend the incidence of these three types of disasters continue to increase from year to year. Therefore, the use of data mining is used to solved this problem. Association rule mining is one of data mining methods that can identity similiarity relationship between items. . The result using arules aprori methode given score > 1 there is indicated that is given a great accuracy can use become a reference for predicting disister. the result givens 69 rules with one of the rules in mounth July with cause information downpour can be occure flood with confidence 0.889 and support 0.012 from all transaction of disaster. And all the lift score on this analysis method given score > 1 there is indicated that is greats accuracy can use become a reference for predicting disister.