The uniform method for the assessment of radon potential of soils (so-
called Radon Risk Classification of Foundation Soils) used in the Czec
h Republic is based on determination of soil-gas radon concentration a
nd on permeability classification. This paper discusses the results of
studies to improve the method based on recent experience. The soil-ga
s radon concentrations may vary, often greatly, over small distances.
An occurrence of heterogenities often indicates the presence of faults
or tectonic zones. Any evaluation based on a single measurement is al
most worthless. The sampling depth of 0.8 m represents a reasonable co
mpromise between the required reduction of errors caused by the radon
variability and the practicability of the method under field condition
s. Observed temporal variations of soil-gas radon concentrations may o
ften correspond not only to real variations of the measured physical q
uality, but also to fluctuations and errors connected with the samplin
g and measuring technique. A reproducibility of the radon risk classif
ication is satisfactory. Neither a normal nor a log-normal model is ge
nerally applicable for a description of soil-gas radon concentration d
ata. The field of soil-gas radon concentration measurement is characte
rized by a relatively high level of knowledge. More problems are conne
cted with the determination of the second main parameter-soil permeabi
lity. Copyright (C) 1996 Elsevier Science Ltd.