Principal component analysis was used to identify the parameters that influence the column-to-column and batch-to-batch reproducibility of retention times and retention factors measured on Symmetry C18, Kromasil C18, Luna C18 (2) and Vydac RP C18, all reversed-phase silica columns. We devised a procedure that allows the determination of the differences in column volume and packing density between two columns, provided that these columns are packed with identical stationary phases (i.e., phases that originate from the same batch). Principal component analysis of the retention times confirmed that the column-to-column variations of the column volume and the total porosity of the bed are the factors that influence the reproducibility of the retention times, the column volume being the major factor. For the fluctuations of the retention factors, the column phase ratios (or the bed porosities) and some specific, secondary retention mechanisms are responsible. All the C18 columns investigated proved to behave in a very similar fashion. Two principal components were always sufficient to characterize the variations of either the retention times or the retention factors.