The objective of this work is to find out how reliably knee joint alignment can be measured from a standardized photograph and what influence changes in the standing position have on the angles measured. The interrater, intrarater, and test-retest reliability were evaluated. The influence of image-object distance, the distance between the legs and leg rotation on the measured angles was evaluated. In addition to the digital photographs, 10 full-length radiographs were obtained in an upright position to determine whether the measured angles represent the anatomic axis or mechanical axis. There was high correlation between the interrater (ICC 0.997), intrarater (ICC 0.989) and test-retest reliability (ICC 0.904). Only slight deviation was found with the changes in radiograph-object distance (0 degrees -1.8 degrees ). With feet together varus malalignment was greater. Leg rotation showed a strong influence on the measured results (ICC 0.658). The angle measured in the digital photographs reflects the mechanical axis with only slight deviation (0.12 degrees -1.9 degrees ). The measurement of the clinical axis using standardized radiography is highly reliable and can be used for individual follow-up of varus and valgus malalignments.