Abstract: Test automation is an important factor for the quality of Android applications. In order to improve the efficiency of automation, test case generation and the ability of abnormity discovering in traditional automated testing of Android application, an automatic test method for Android application was proposed based on particle swarm optimization algorithm, named particle swarm optimization for Android (PSOA). The PSOA method was arranged with three functions, including information collecting of control units, test process modularizing, test case generating and test data optimizing with the particle swarm algorithm. Experimental comparing with traditional automated test, the proposed method can successfully improve the ability of abnormity discovering, the efficiency of test case generation and the rate of automation.