1. The purpose of the study was to evaluate a non-invasive technique for measurement of microvascular permeability to a small hydrophilic solute. 2. The technique measures the clearance of 99(m)Tc-labelled diethylenetriaminepenta-acetic acid (99(m)Tc-DTPA) from plasma into interstitial fluid in a limb after intravenous injection and uses a scintillation probe and a technique of graphical analysis called the Patlak plot, the uptake constant of which reflects 99(m)Tc-DTPA transfer from plasma to interstitial fluid, Using deconvolution analysis, the retention function in the limb of intravenous 99(m)Tc-DTPA was also measured. 3. The clearance values given by these two analytical techniques were compared with clearance from the same vascular bed after bolus femoral intra-arterial injection of 99(m)Tc-DTPA. 4. Sixteen patients undergoing routine diagnostic arteriography were studied: six received sequential femoral intra-arterial injections of 99(m)Tc-labelled human serum albumin (HSA) and 99(m)Tc-DTPA, two received sequential intra-arterial and intravenous injections of 99(m)Tc-HSA and eight received sequential intra-arterial and intravenous injections of 99(m)Tc-DTPA. Tissue uptake and clearance were recorded from the limb with a scintillation probe and plasma clearance by arterial blood sampling. Tracer recirculation was addressed using a second scintillation probe over the contralateral limb. 5. After intra-arterial injection, 99(m)Tc-HSA clearance was monoexponential, reflecting intravascular transit, and was completed by 2-5 min in seven subjects and in about 10 min in one, The corresponding 99(m)TcDTPA clearance curves in the six subjects who also received intra-arterial DTPA were biexponential, analysis of which yielded a 99(m)Tc-DTPA extraction fraction of about 0.6, By comparison with 99(m)Tc-HSA clearance, the first exponential clearly corresponded to intravascular transit of unextracted 99(m)Tc-DTPA. 6. In the eight patients given sequential intra-arterial and intravenous injections of 99(m)Tc-DTPA, the second exponential recorded after intra-arterial injection, representing 99(m)Tc-DTPA clearance from the interstitial fluid, agreed well with (a) the Patlak uptake constant recorded over the limb after intravenous injection, representing clearance from plasma into the interstitial fluid and (b) the retention function of 99(m)Tc-DTPA in a limb calculated by deconvolution analysis. The mean clearance following intraarterial injection (expressed in relation to extracellular fluid volume) was 9.6 (SD 2.4) mlmin(-1)100ml(-1), while the corresponding mean clearance after intravenous injection was 8.8 (2.1) ml min(-1) 100ml(-1) calculated by Patlak analysis and 10.5 (2.7) ml min(-1) 100 ml(-1) by deconvolution analysis. 7. We conclude that, under the conditions of measurement, 99(m)Tc-DTPA is about 60% extracted into the interstitial fluid in a single pass through an extremity and that clearance into the extravascular space can be measured with reasonable accuracy after intravenous injection.