We read with interest this report by Bagiotti et al. on colour Doppler traces from the main testicular artery. We recently conducted a similar study comparing testicular arterial blood flow in normal subjects and unobstructive azoospermic men. We also measured the Doppler arterial blood flow variables from three specific areas, i.e. supratesticular, capsular and intratesticular, as described previously . There was a very significant difference in peak systolic velocity (P < 0.01) and resistance index (P < 0.001) between the groups, similar to that in the published study. The intratesticular arteries often had very little flow and it was therefore technically difficult to measure. The supratesticular arteries were more variable in all the recorded blood flow variables, thought to be related to the possibility of erroneous measurements of arterial vessels that may anastomose with the testicular artery, i.e. the cremasteric, vasal and epididymal arteries. The capsular arteries, which run closely around the testis, are easily detected on the periphery of the grey-scale image of the testis using colour flow Doppler. The velocity of blood flow in the capsular arteries is greater than in the intratesticular arteries, and the assessment is less prone to artefact. Most importantly, the flow variables recorded from these capsular arteries showed the most significant correlation with spermatogenesis. In addition to conventional means, we recommend estimating arterial flow within the capsular arteries of the testes when assessing azoospermic men, as a simple and noninvasive tool to aid the differentiation between those with primary spermatogenic failure and obstructive causes. More extensive use of Doppler ultrasonography in the infertile man may also allow the prediction of successful sperm retrieval techniques.