In the course of a routine thermoluminescent dating program the uranium and thorium concentrations were measured in 39 samples by fission track counting and neutron activation analysis to evaluate the annual dose-rate. These values show large discrepancies compared with these obtained from alpha counting. This can only in part be explained by radiochemical disequilibrium. It rather seems that overcounting due to radon escape is the reason for those discrepancies. It is concluded that dose-rate evaluation even of the alpha contribution by alpha counting is liable to large errors. (See also <b>AATA</b> <i>18-64</i> and <i>20-1487</i>) -- AATA