The distribution of ferric iron (Fe3+) between the octahedral and tetrahedral sheets of smectites is still an active problem due to the difficulty of identifying and quantifying the tetrahedral ferric iron (Fe3+). Mössbauer spectroscopy has often been used to address this problem, with the spectra being fitted by a sum of doublets, but the empirical attribution of each doublet has failed to yield a uniform interpretation of the spectra of natural reference Fe3+-rich smectites, especially with regard to Fe3+, because little consensus exists as to the Fe3+ content of natural samples. In an effort to resolve this problem, the current study was undertaken using a series of synthetic nontronites [Si4–xFex3+] Fe23+O10(OH)2Nax with x ranging from 0.51 to 1.3. Mössbauer spectra were obtained at 298, 77, and 4 K. Statistically acceptable deconvolutions of the Mössbauer spectra at 298 and 77 K were used to develop a model of the distribution of tetrahedral substitutions, taking into account: (1) the Fe3+ content; (2) the three possible tetrahedral cationic environments around Fe3+, i.e., [4Si]-(3Fe3+), [3Si Fe3+]-(3Fe3+), and [2Si 2Fe3+]-(3Fe3+); and (3) the local environment around a Fe3+, i.e., [3Si]-(2Fe3+) respecting Lowenstein's Rule. This approach allowed the range of Mössbauer parameters for Fe3+ and Fe3+ to be determined and then applied to spectra of natural Fe3+-rich smectites. Results revealed the necessity of taking into account the distribution of tetrahedral cations (R3+) around Fe3+ cations to deconvolute the Mössbauer spectra, and also highlighted the influence of sample crystallinity on Mössbauer parameters.