The sparkle/AM1 model for the quantum chemical prediction of coordination polyhedron crystallographic geometries from isolated lanthanide complex ion calculations, defined recently for Eu(III), Gd(III), and Tb(III) (Inorg. Chem. 2005, 44, 3299) is now extended to La(III) and Lu(III). Thus, for each of the metal ions we chose a training set of 15 complexes that possess various representative ligands of high crystallographic quality (R factor < 0.05 angstrom) and oxygen and/or nitrogen as coordinating atoms. In the validation procedure we used a set of 60 more La(III) coordination compound structures, as well as 15 more Lu(III) coordination compound structures, all of high crystallographic quality. For both the 75 La(III) compounds and the 30 Lu(III) compounds, the Sparkle/AM1 unsigned mean error, for all interatomic distances between the metal ions and the ligand atoms of the first sphere of coordination, is 0.08 angstrom, thus comparable to the accuracy normally achievable by present day ab initio/ECP calculations, while being hundreds of times faster.