We propose an improved equation for calculating the uranium concentration in thick samples based on induced fission-track counts in an external detector that takes into consideration (1) the fission-fragment ranges in the sample and external detector, (2) the etchable track length and (3) the track counting efficiency in the external detector. The values of these parameters have been determined by calculation and experiment and are shown to have a significant effect on the calculated uranium concentrations. The new equation was tested by calculating the uranium concentrations in standard uranium glasses (CN-5; IRMM-540R) and apatite samples (Durango; horse tooth) in which the uranium content was also determined with independent methods (INAA; ENAA; TIMS). The results show that: (1) accurate measurements with the fission-track method are feasible within a broad range of uranium concentrations and (2) uranium determinations based on standards are only accurate if the standard and sample are made of the same material. Because the absolute fission-tack dating method is also based on accurate thermal neutron fluence measurements and similar correction factors for the track registration and counting efficiencies, this study provides a strong endorsement for the fact that absolute fission-track ages are accurate.