Methods of quantifying changes in spectrum analysed Doppler waveforms have been assessed in an experimental canine model. Artificial stenoses of varying (0-95%) area reduction have been inserted in the iliac artery of five anaesthetised greyhounds. 133 sets of Doppler waveforms recorded distally have been subjected to principal component analysis (PCA), transfer function analysis (TFA) and pulsatility index calculation (PI). Transfer function damping factor delta and PI were only able to distinguish those stenoses of greater than 85% area reduction. PCA, however, was more reliable in this range, and was usually able to distinguish stenoses of 65% and 77% area reduction. We conclude that PCA deserves further assessment in patients with aorto-iliac disease.