Deadtime count loss may cause significant errors in quantitative measurements performed by scintigraphic techniques. Deadtime depends on the scattering condition, and may vary among patients. The aim of this study was to estimate deadtime in imaging a human subject. A cylindrical phantom filled with various concentrations of technetium-99m solution was imaged with a gamma camera, and deadtime was assessed assuming the gamma camera to be a non-paralysable system (multi-dose method). Deadtime for the cylindrical phantom was also estimated using a point source to monitor counting efficiency (reference source method). Radionuclide angiography of the chest was performed in 38 patients with a reference source in the field of view, and temporal changes in count loss were assessed. The deadtime for each patient was calculated by the reference source method. The deadtime for the cylindrical phantom was estimated as 6.96+/-0.09 micros by the reference source method and was almost identical to that obtained by the multi-dose method (6.93 micros). Deadtime ranged from 6.01 to 9.58 micros in radionuclide angiography, and was positively correlated with the ratio of body weight in kg to body height in cm (y=8.566 x+5.611, r=0.869). Count loss was successfully corrected using the deadtime predicted with the regression line. In summary, the deadtime in imaging a human subject has a wide range and is related to the constitution of the subject. Correction for count loss using a deadtime predicted for each patient may be helpful in improving the reliability of quantitative nuclear medicine.