X-ray microtomography data of iron ore green pellets of approx. 12 mm in diameter were recorded using a commercial instrument. The reconstructed volume after thresholding represented a unique dataset consisting of a three-dimensional distribution of equiaxed objects corresponding to bubble cavities. This dataset was used to successfully validate a stereological method to determine the size distribution of spherical objects dispersed in a volume. This was achieved by investigating only a few cross-sectional images of this volume and measuring the profiles left by these objects in the cross-sectional images. Excellent agreement was observed between the size distribution of the bubble cavities obtained by directly classifying their size in the reconstructed volume and that estimated by applying the aforementioned stereological method to eight cross-sectional images of the reconstructed volume. Subsequently, we discuss the possibility of calibrating X-ray tomography data quantitatively using the size distribution of the bubble cavities as a figure of merit and the results obtained by applying the stereological method to SEM images as reference data. This was justified by considering the validity of the stereological method demonstrated by tomography, the accurate thresholding made possible by back-scattered electron imaging and the solid reproducibility of the results obtained by SEM. Using different threshold values for binarization of the X-ray microtomography data and comparing the results to those obtained by SEM, we found that X-ray microtomography can be used after proper calibration against SEM data to measure the total porosity of the bubble cavities but can only provide a rough estimate of the median diameter because of the limited resolution achieved in this study. (C) 2012 Elsevier B.V. All rights reserved.