A computer-based version of the rod-and-frame test for verticality perception has been developed. The task tested verticality perception of subjects by using the computer mouse to rotate a bright rod to a vertical orientation within a square frame that was tilted through a range of angles from 30 degrees counterclockwise to 30 degrees clockwise. In 16 healthy subjects, errors from the true vertical were generated in the direction of the frame tilt. The errors reached statistical significance ( P <.05) when the frame was rotated by 10 degrees in either direction. In a second study, the images were presented to the subjects who wore a pair of video eyeglasses. The pattern of errors generated by rotation of the frame were similar but were of larger amplitude than those produced in the first experiment and reached statistical significance with frame angles of 10 degrees and 20 degrees in both directions. These preliminary investigations suggest that the computer-based rod-and-frame test may provide a more convenient alternative to the mechanical rod-and-frame test for verticality perception.