Multiple treatment options and risk assessment in cerebrovascular diseases are the actual challenges in diagnostic as well as in interventional neuroradiology.Acute ischemic stroke essentially requires rapid detection of the location and extent of infarction and tissue at risk for making treatment decisions. In the acute setting, modern multiparametric perfusion imaging protocols help to determine infarct core and adjacent penumbral tissue, and they enable the estimation of collateral flow of intra- and extracranial arteries. In subacute delayed cerebral ischemia (DCI) after subarachnoid hemorrhage (SAH) or chronic occlusive neurovascular diseases estimation of residual and collateral flow may be even more difficult.Prediction of sufficient or insufficient supply of brain tissue may be essential to balance conservative against interventional therapies. However, so far no established reliable thresholds are available for determining tissue at acute, subacute, chronic progressive, or chronic risk.Reliable and reproducible thresholds require quantitative perfusion measurements with a calibrated instrument. But the measurement instrument is not at all defined-a variety of parameter settings, different algorithms based on multiple assumptions and a wide variety of published normal and pathologic values for perfusion parameters indicate the problem. In the following text, we explain how deep the problem may be enrooted within techniques and algorithms impeding broad use of perfusion for many clinical issues.