The efficiency calibration of whole-body counters (WBCs) for monitoring of internal contaminations is usually performed with anthropomorphic physical phantoms assuming homogeneous activity distribution. Besides the inherent limitations of these phantoms in resembling the human anatomy, they do not represent a realistic activity distribution, since in real situations each incorporated radionuclide has its particular biodistribution after entering the systemic circulation. Moreover, the activity content in the different organs and tissues comprising the biokinetics is time dependent. This work aims at assessing the whole-body counting efficiency deviations arising from considering a detailed voxel phantom instead of a standard physical phantom (BOMAB) and at evaluating the effect of the anatomical differences between both phantoms. It also aims at studying the efficiency considering the biodistribution of a set of radionuclides of interest incorporated in the scope of environmental and occupational exposures (inhalation and ingestion) and at computing the time-dependent efficiency correction factors to account for the biodistribution variation over time. For the purpose, Monte Carlo (MC) simulations were performed to simulate the whole-body counting efficiencies and biokinetic models were used to estimate the radionuclides' biokinetic behaviour in the human body after intake. The comparison between the efficiencies obtained with BOMAB and the voxel phantom showed deviations between 1.8 and 11.7 %, proving the adequacy of the BOMAB for WBC calibration. The obtained correction factors show that the effect of the biodistribution in the whole-body counting efficiency is more pronounced in cases of acute activity uptake and long-term retention in certain organs than in cases of homogeneous distribution in body tissues, for which the biokinetics influence can be neglected. This work further proves the powerful combination of MC simulation methods using voxel phantoms and biokinetic models for internal dosimetry studies.