Results are repoerted for238U,234U,232Th, and230Th determinations in 19 rock samples from a uranium mine, performed independently, byb three different laboratories. Uranium and thorium isotopoic activities were determkined by alpha spectrometry, after different pre-concentration and counting sample preparation techniques., Additionally, total concentrations of uranium were determined by fluorimetry and gamma spectreometry. the folloing conclusions could be drawn from this intercomparison test: (1) The results for238U specific activity agreed with the amjority of results within 10%. Lincar correlation coefficients between the three data sets were 0.999. However, for a few samples of much higher uranium concentrations, large deviations were observed, indicating problems of, sample heterogencity. (2) For the234U/238U activity ratio data, a still closer agreement was obtained (5%), as computation of the activity ratios did not, require information on the yield of the used tracer spike (232U). (3) The results for232Th specific activities and230Th/234U activity ratios showed larger deviastions between the three laboratories (typically up to 15%, in some cases still ore). Different Th-isotopes (228Th,234Th and229Th) have been used as yield tracers. The data indicates, however, that the observed deviations are not simply a consequence of a systematic difference in the calibration of the different spikes, but, probably cased by other errors such as incomplete sample dissolution, sample heterogencity, tec. The limitations of alpha spectrometry will be discussed and an application of the developed methods shown.