Environmental samples from a wide-range of aquatic and soil deposits, mainly of Scandinavian origin, were analysed for Th-234 (U-238) using low-level gamma-spectrometry. The diversity of the samples, in terms of composition and ages, allowed a detailed evaluation of the analytical problems associated with gamma-ray spectrometry with focus on the reliability of the Th-234 peaks for absolute determination of the Th-234 activities. The X-ray contributions in the 93 keV peak were compared with the corresponding self-absorption corrected activities of the 63 keV peak. These X-ray contributions were, also, correlated with the U-238, Th-232, U-235, K-40 and Cs-137 activities of the samples. Despite the difficulties imposed by the self-absorption corrections, the 63 keV peak is still the best option. Large variability in the 93 keV peak interferences, due to X-rays from Th, exists in sediment and soil samples. Only in the case of young ombrotrophic peat samples was it possible to conclude that the 93 keV peak is free from X-ray contributions and can be as good as the 63 keV Monte-Carlo self-absorption corrected peak. X-ray contributions in the samples correlated with the 238 U and 232 Th activities, only, in closed environmental systems where a secular equilibrium with the daughters of the U/Th series can occur.