A surgical stereoscopic visualization system having a microscope head, dual cameras mounted on or proximate the microscope head for real time video capture of a subject in an operative field, a video processing and display system to receive separate video streams from the cameras and to combine the separate video streams into a time-aligned stereoscopic image. The processed image is displayed on a matrix display screen in at least one movable head-mounted stereoscopic display. Movement of the head mounted display is accomplished with a support structure on a headband that accurately positions the display screen in front of the user's eyes, yet allows the user to precisely move the display screen either manually or by motorized means into and out of the user's line of sight, thereby facilitating uninterrupted use of the hands while working in the surgical suite.