A rigid-ion lattice-dynamical treatment accounts reasonably well for most of the atomic displacement parameters (ADPs) reported in the literature for a synthetic sample of spessartine Mn(3)Al(2)Si(3)O(12); the model accounts less well for the ADPs of the Mn atom. These results are similar to those obtained for the corresponding garnets pyrope Mg(3)Al(2)Si(3)O(12) and almandine Fe(3)Al(2)Si(3)O(12). The results are confirmed by a very good agreement of the lattice-dynamical estimates with the experimental values of the heat capacity that are currently available, as well as those of the Raman and infrared vibrational spectra. The anomalous behaviour of the Mn atom is reflected in the entropy values. As is the case for pyrope and almandine, here the agreement with the experimental results can be markedly improved if the existence of a low-temperature order-disorder transition concerning the Mn atom is considered. The lattice-dynamical estimates for all the atoms presented here do not agree with the corresponding ADPs obtained in a crystal structure refinement of a supposedly 'pure' natural sample from Elba. In turn, these experimental ADPs do not agree with those observed for the synthetic sample, thereby showing a physical difference; such a possibility is in agreement with recent observations by other authors, who found an unexpected fluorine content in several natural specimens of spessartine. Therefore, theoretical estimates of ADPs are reliable enough to point out unexpected situations if disagreement with the observed values occurs.