A deadtime correction method is proposed for quantitative measurements in positron emission tomography. The correction is based on the observation that a deadtime loss of the random coincidence events corresponds to that of the single events and of the total coincidence events. Using the proposed correction method, the deadtime loss was kept within 1% up to the true coincidence rate of 50 X 10(3) cps per plane for a cylindrical phantom of 20 cm in diameter. Accuracy of the method is confirmed to be independent of the size of the object.