AbstractBecause of the large number of stimulus configurations that may be displayed on a computer controlled color monitor, direct measurement of the relation between the digital input values to the frame buffer and the output of the monitor is not possible. To calibrate a monitor, it is necessary to make assumptions about the monitor's performance. These assumptions allow the monitor's output for any set of input values to be predicted from the input values and the results of a feasible number of calibration measurements. This article describes a set of assumptions that are sufficient to render the problem of monitor calibration tractable. These are the assumptions of phosphor constancy, phosphor independence, spatial independence, and the single scale factor assumption. If these assumptions hold, a monitor may be calibrated by measurements of the phosphor spectral power distributions, measurements of the phosphor input‐output relations, and measurements of the spatial scale factor at a number of locations. Violations of these assumptions limit the accuracy of calibration. The results of measurements that test the validity of these assumptions for a Barco 5351 monitor are presente