Traditionally, Doppler ultrasound has been used to estimate blood flow as the mean velocity multiplied by the vessel area, but this is subject to significant errors and may be difficult to perform accurately. Microbubbles, developed as contrast agents for ultrasound, were initially envisaged as useful for increasing the intensity of echoes and thus rescuing Doppler studies that were technical failures because of attenuated signals or very slow flow. However, they can act as tracers and, by analogy with isotope techniques, can be used to measure blood flow with transit-time methods which exploit both arterial and venous time-intensity data. An acceptable compromise is to acquire both a tissue intensity curve and one from the feeding artery. The transit of microbubbles across an organ or tissue can be used to estimate haemodynamic alterations, e.g. the arterialisation of the supply to the liver in malignancies and cirrhosis and the delayed arterio-venous transit in the transplant kidney during rejection. The fragility of microbubbles can be turned to advantage by being exploited to create a negative bolus by exposing a tissue slice to a high power beam. The rate of refilling of this slice by circulating microbubbles can then be followed with a low-intensity monitoring beam and the resulting rising exponential curve analysed to extract indices of both the reperfusion rate (the slope) and the fractional vascular volume (the asymptote). The product of these is a measure of true tissue perfusion.