Phono Arthrometry is the study of sound naturally generated by a joint in motion, usually associated with the knee. The technique has not achieved use in front line health care, though has received modest attention in the literature since its first appearance in 1902 with a total of 52 publications identified by this thesis. The Phono Arthrometric traces are prone to extreme levels of variability, a difficulty recognised by some of the literature and a plausible reason for the technique not being fully developed. This work seeks to Identify the causes of the intense variability and attempt to account for them using Multi Dimensional Attribute Analysis, a little known data processing technique.