Abstract Digital image distance measurements have two sources of error: they are the inherent image pixelation error and modality-specific registration errors. A theoretical analysis was performed to determine the pixelation error uncertainty in ultrasound (US) clinical and phantom digital image distance measurements. For horizontal or vertical phantom image distance measurements, the average absolute image pixelation error was found to be 0.5 pixel width. In the clinical horizontal case, the combined absolute distance measurement error due to pixel size and operator cursor placement errors was found to be 1.5 pixel widths (for a cursor placement error of 1 pixel width). For image clinical distance measurements, a formula is given for the dependence of the measurement error on the inclination of the measured structure in the pixel matrix. A protocol is recommended that minimizes all sources of error in clinical image distance measurements.