Audio and video devices can be virtualized using synchronous A/V streaming. When a video device such as a webcam is connected to a client terminal while the client terminal has established a remote session on a server, video data generated by the video device can be encoded into an encoded video stream using the H.264 or similar standard. Additionally, audio data corresponding to the video data can be embedded into the encoded video stream. The encoded video stream with the embedded audio can then be transferred to the server. The server-side components can extract the audio from the stream and perform the necessary decoding on the video and possibly on the audio data prior to providing the decoded data to drivers that function as virtualized audio and video devices on the server.