One of the basic steps in luminescence dating is the determination of the annual radiation dose. Among the suitable determination methods in the case of sediment dating is HPGe gamma-ray spectrometry, primarily yielding the concentrations of K, Th and U via measurement of gamma-rays emitted by 40K and by 232Th and 235,238U and their decay products. These determinations involve both large-volume samples (with cylindrical or Marinelli geometry) and low-energy gamma-rays (down to the 210Pb 46.5 keV line). In view of this, calculations have been made in the present work in order to investigate possible inaccuracies related to gamma-ray attenuation, which may be different in samples and calibrants which have considerably different composition (elemental concentration and packing density). The calculations are based on the introduction of correction factors containing “effective solid angles” (proportional to the peak detection efficiencies), which are believed to give “correct” concentration results. It was found that, in some cases, significant errors could be committed when not performing proper corrections. Therefore, it is concluded that in practice this possible source of inaccuracy should be kept in mind, thus making it necessary to have a fair knowledge of the properties of the materials (samples and calibrants) under investigation.