TK R.A UPEKA II is one of the play schools located in North Jakarta. TK R.A UPEKA II has problems in detecting personality disorders of each student, due to the high cost of calling a psychiatrist to find out the disorders experienced by students. One of the disorders that often occurs in kindergarten students is Attention Deficit Hyperactivity Disorder (ADHD). Attention Deficit Hyperactivity Disorder (ADHD) is a developmental disorder that increases motor activity that causes children to behave excessively or outside normal limits. To overcome this problem, the author conducts research to create an expert system application by applying the Dempster Shafer method as a way of calculating the confidence value of a diagnosis. This calculation is done by comparing each belief value of the 2 initial symptoms that occur in children and then compared with the belief values of other symptoms. This application aims to help TK R.A UPEKA II to make it easier to quickly detect Attention Deficit Hyperactivity Disorder (ADHD) like an expert or a psychiatrist. In this writing, the author uses the Java programming language and Hypertext Preprocessor (PHP), and uses the MySQL database as a storage medium for the data. By using the Waterfall software development method in making the design flow. It is hoped that the application of the expert system with the dempster shafer method can make it easier to detect Attention Deficit Hyperactivity Disorder (ADHD) in students at TK R.A UPEKA II, so that students who are diagnosed can be treated medically. And also helps the school in assessing the behavior of its students.