The application of a detection system using the YOLOv5 algorithm is very important for current parking management with both open and closed parking types. In parking management, there are often problems that cause less effective parking lot management. This purpose research aims to manage the availability of parking lots more effectively by classifying parking slots and displaying rectangles with different color labels and displaying the F1 Convidence Curve to see the accuracy value of the model to be ready for detection. In this research, the model used is YOLOv5 which can detect accurately and quickly using the Python programming language. The model is trained and evaluated and tested to produce a model that has the expected level of accuracy. The detection results display the parking lot availability classification in the form of purple and yellow rectangles. The purple color means that the parking space is occupied and the yellow color means that the parking space is still empty. The detection results are also displayed in real time so that the detection system can be used immediately and continuously. The F1 Convidence value shows 0.87 in all classes so that the model can be considered to have high accuracy in detecting objects in the parking lot.