Background and purposeAs a non-invasive, side effect-free and cost-effective method, ultrasonography represents the method of choice for the diagnosis of renal artery stenosis. Four different criteria in total, including two direct criteria in peak systolic velocity (PSV) and renal aortic ratio (RAR) and two indirect criteria in resistance index (RI) and acceleration time (AT) for the measurement of relevant renal artery stenosis are described, each demonstrating highly variable accuracy in studies. Furthermore, there is controversy over the degree beyond which stenosis becomes therapeutically relevant and which ultrasound PSV is diagnostically relevant in terms of stenosis grading.Material and methodsThis article gives a critical review based on a selective literature search on measurement methodology and the validity of ultrasound in renal artery stenosis. A critical evaluation of methods and a presentation of measurement principles to establish the most precise measurement method possible compared with the gold standard angiography, as well as an evaluation of the importance of computed tomography angiography (CTA) and magnetic resonance angiography (MRA).Results and conclusionsThe PSV provides high sensitivity and specificity as a direct measurement method in stenosis detection and grading. Most studies found sensitivities and specificities of 85–90 % for > 50 % stenosis at a PSV > 180–200 cm/s in ROC curve analysis. Other methods, such as the ratio of the PSV in the aorta to the PSV in the renal artery (RAR) or indirect criteria, such as side to side differences in RI (dRI) or AT can be additionally used to improve accuracy. Contrast-enhanced ultrasound improves accuracy by means of echo contrast enhancement. Although in the past only high-grade stenosis was considered relevant for treatment, a drop in pressure of > 20 mmHg in > 50 % stenosis (PSV 180 cm/s) is classified as relevant for increased renin secretion. Stenosis in fibromuscular dysplasia can be reliably graded according to the continuity equation. Although the available studies on the grading of in-stent restenosis are the subject of controversy, there is a tendency to assume higher cut-off values for PSV and RAR. Whilst MRA and CTA demonstrate an accuracy of > 90 %, this is at the cost of possible side effects for patients, particularly in the case of pre-existing renal parenchymal damage.Additional online materialThis article includes two additional video sequences on visualizing renal artery stenosis. This supplemental material can be found under: dx.doi.org/10.1007/s00772-015-0060-3