In routine thermoluminescence (TL) dating, there is often a tendency to omit an absolute determination of the thorium and uranium content in the sample and simply measure the total alpha count rate. The annual dose is then calculated assuming equal activities for a Th/U weight ratio of about 3.17. In fact this ratio is varied significantly in pottery. Calculation of the annual dose conversion factors from data using the total alpha count rate in a selected range of Th/U weight ratios shows significant differences. The results indicate that the additional error contribution to the final TL age is in the range ±2.5% to ±4.1% for most cases where the Th/U ratio lies between 1.1 and 9.5.