A wide range of collision cross section (CCS) databases for different families of compounds have recently been established from ion mobility mass spectrometry (IM-MS) measurements. Nevertheless, the need to validate these new data sets to provide the necessary confidence about the use of this parameter is increasingly expressed by the scientific community. If such a validation requires that complementary mass spectrometry experiments are conducted, it also appears that alternative strategies can contribute to the validation of such empirical data. In particular, in silico approaches are relevant to compute theoretical CCS values, to be compared to experimental ones. A recently published CCS database for 300 steroids allowed one to observe experimentally significant deviations of the expected CCS versus m/z correlations for some compounds. The present work attempts to rationalize such deviations with Density Functional Theory (DFT) calculations. MNIS/6-311++G(d,p) investigations have been carried out, starting with a conformational analysis of a sample of 20 selected steroids and the determination of their preferred gas-phase ionization site. CCS values were then computed and compared to the experimental data. This approach allowed one to rationalize the experimental trends, providing an accurate description of the key properties of the various steroids considered.