This study aims to create a smart parking system that can find empty parking slots using haar cascade classifier algorithm through an application on a smartphone. The smart parking system is a system created to find out empty parking slots through an application on smartphones. The smart parking system uses a web camera that is connected to an application on a smartphone with implement the IoT system. Using the Haar Cascade Classifier algorithm, a web camera installed in a parking lot can detect parking symbols. If the camera detects a parking symbol, the system reads that the parking slot is empty. Conversely, if there is a car that covers the symbol and the symbol is not detected by the camera, then the system reads that the slot parking is filled. Testing the accuracy value is carried out using the confusion matrix method from four trials get the highest accuracy value 96.29%, the highest accuracy value is cascades training with 108 images and use 15 stages.