Abstract: Foam metal is a kind of new material that can inhibit gas explosion flame and pressure wave, and there are many factors that affect the effect of suppression and isolation, and the testing process is expensive and the period is long. We use artificial neural network （ＡＮＮ） method to predict suppression and isolation effect of foam metal to reduce the material loss in experimental period, optimize parameters combination and improve the efficiency of the experiment. The results show that the BP neural network can be used for explosion suppression and isolation effect prediction of foam metal. The optimal network can predict the biggest attenuation of pressure and temperature, and the average error can reach 13% and 4% respectively when BP network has 10 neurons and its transfer function uses “logsig”, “purelin”. The research results show that artificial neural network can be used for suppression and isolation effect prediction of foam metal.