The lipophilic 99mTc-D,L-hexamethylene-propyleneamine oxime (99mTc-HMPAO) has been developed for regional cerebral blood flow (rCBF) measurements by single photon emission computed tomography (SPECT). The molecule is unstable and converts rapidly from the lipophilic form, which passes the blood-brain barrier (BBB), to the hydrophilic form, which is unable to pass the BBB and is trapped in the brain. The rate-limiting step for this conversion is probably dependent on the reductant gluthathione. The lipophilic input to the brain can be estimated by rapid octanol extraction of lipophilic tracer from arterial blood. The input takes place during the first few minutes after tracer injection. The first-pass extraction from blood to brain E is high (0.72 at a CBF of 0.59 ml/g/min) in human studies as measured by the indicator dilution method. It is dependent on the CBF level and decreases when CBF increases. It is also dependent on binding to proteins and blood constituents. 99mTc-HMPAO is initially distributed like rCBF. In measuring the retention in the human brain after intravenous and intracarotid injection of 99mTc-HMPAO, an early back-diffusion (brain to blood) is seen. This lasts only 2-3 min. The back-diffusion is flow dependent, leading to a preferential loss of activity from the high flow regions of the brain. This can be corrected for by an algorithm. The effect of the algorithm is that the steady-state 99mTc-HMPAO distribution images obtained from 10 min to 2-3 h after injection of tracer agrees more closely with rCBF images as measured by reference CBF methods using SPECT and positron emission tomography (human studies) and quantitated autoradiography (rats). The retention in the brain is very stable when the early back-diffusion has ceased, and only a small loss of tracer amounting to 0.4%/h is observed in most human cases during the next 24 h. This review concludes that 99mTc-HMPAO is suitable for measurements of rCBF by SPECT. A few examples of clinical application are given.