Purpose: A new method for examining and measuring the weight-bearing knee in computed radiography (CR) and picture archiving and communication systems (PACS) has recently been developed on examination equipment used in QUESTOR Precision Radiography (QPR). QPR is a method for the standardised examination of the knee, and generates 9 angles (e.g. hip-knee-ankle (HKA) angle) and 10 distances, corrected for parallax and magnification. The aim of this present study was to evaluate the reproducibility of this newly developed method and to determine intra- and interobserver variation in its measurements. Material and Methods: The images were generated on a CR system, archived, and transferred to a multimodality work-station that had a monitor with a resolution of 1280x1024 (1kx1k). Photostimulable phosphor plates were used to generate images with a matrix of 1760x2140 pixels (pixel size 0.2 mm, corresponding resolution 2.5 1p/mm). Ten volunteers without knee problems were examined twice, bilaterally, by 2 different radiology technologists. Results: The total reproducibility of the method was good, offering an HKA reproducibility of ±2.64± in slight flexion and ± 1.62± in extension, at 95% confidence. The intra— and inter-observer variations were low with a reduction of the intra-observer variations in all measurements (except one) by a factor of 2.8 on average compared with the original QPR method for conventional film-screen radiography. Conclusion: When QPR is modified for CR, it fulfils the requirement of reproducibility in measurements of the weight-bearing knee.