Surveillance cameras generally have a wide field of view and do not capture the human face with a resolution that is sufficient for machine recognition. To overcome this problem, this paper presents a realtime face detection and tracking algorithm using a single PTZ (Pan, Tilt, Zoom) camera. Unlike existing algorithms which track the human face in the field of view of a fixed camera, the proposed algorithm adjusts the pan, tilt and zoom of a PTZ camera in real-time so that human faces can be acquired at the camerapsilas maximum resolution. The proposed algorithm addresses challenges like network delays, camera movement lags, oscillations and lost faces. Experiments were performed in realistic surveillance scenarios and accurate real-time tracking with pan, tilt and zoom was achieved.