An audio-video synchronization detection device and a method thereof are provided. The audio-video synchronization detection device detects an object under test which is capable of generating an image signal and an audio signal. The audio-video synchronization detection device includes a delay circuit, an optical sensor, and a signal processor. The delay circuit delays the audio signal for a predetermined time and generates an audio correction signal accordingly. The optical sensor senses a light emitted by a display panel when the display panel displays the image signal and generates an image sensing signal accordingly. The signal processor calculates a delay time between the audio correction signal and the image sensing signal to obtain a synchronization state between the image signal and the audio signal.