Assessment of cavernosal perfusion in men with erectile dysfunction (ED) relies on Doppler spectrum analysis of pharmachologically stimulated peak systolic velocity (sPSV) in cavernous arteries but its accuracy in identifying men affected by a cavernosal perfusion disorder correlated with atherosclerosis remains undefined. We estimated by B-mode ultrasound, the accuracy of sPSV of cavernous arteries to identify ED with an expected cavernosal perfusion disorder. This was predicted by an elevated intima-media thickness (IMT) of common carotid arteries, a reference methodology to estimate the degree of generalized atherosclerosis, in men exposed to vascular risk factors (VRFs). sPSV and IMT were evaluated in 269 men with ED, 49 had no VRFs, 100 were overweight with/or without hyperlipidaemia, 120 were affected by type 2 diabetes and/or essential hypertension. sPSV was significantly lower (p < 0.05) in patients with VRFs associated with atherosclerosis (IMT ≥ 1 mm) (n = 39) than in men with no VRFs and no atherosclerosis (n = 49). sPSV correlated negatively with age (p < 0.0001), with serum% of glycated haemoglobin (p = 0.010) and with carotid artery IMT (p = 0.013). An sPSV ≤ 30 cm/sec, the cut-off value which showed at receiver operating characteristic curve analysis the combined highest value of sensitivity and specificity, correctly identified only 57% of men in whom a cavernosal perfusion disorder was expected based on the presence of carotid artery IMT ≥ 1 mm combined to the exposure to VRFs. The ultrasonographic evaluation of sPSV had a very limited accuracy in discriminating ED with an expected cavernosal perfusion disorder, based on the presence of a generalized atherosclerosis in men with VRFs.