Materials with apatite crystal structure have a great potential for incorporating the long-lived radioactive iodine isotope (129 I) in the form of iodide (I −) from nuclear waste streams. Because of its durability and potentially high iodine content, the apatite waste form can reduce iodine release rate and minimize the waste volume. Crystal structure and composition of apatite (A 5 (XO 4) 3 Z) was investigated for iodide incorporation into the channel of the structure using Artificial Neural Network. A total of 86 experimentally determined apatite crystal structures of different compositions were compiled from literature, and 44 of them were used to train the networks and 42 were used to test the performance of the trained networks. The results show that the performances of the networks are satisfactory for predictions of unit cell parameters a and c and channel size of the structure. The trained and tested networks were then used to predict unknown compositions of apatite that incorporates iodide. With a crystal chemistry consideration, chemical compositions that lead to matching the size of the structural channel to the size of iodide were then predicted to be able to incorporate iodide in the structural channel. The calculations suggest that combinations of A site cations of Ag + , K + , Sr 2+ , Pb 2+ , Ba 2+ , and Cs + , and X site cations, mostly formed tetrahedron, of Mn 5+ , As 5+ , Cr 5+ , V 5+ , Mo 5+ , Si 4+ , Ge 4+ , and Re 7+ are possible apatite compositions that are able to incorporate iodide. The charge balance of different apatite compositions can be achieved by multiple substitutions at a single site or coupled substitutions at both A and X sites. The results give important clues for designing experiments to synthesize new apatite compositions and also provide a fundamental understanding how iodide is incorporated in the apatite structure. This understanding can provide important insights for apatite waste forms design by optimizing the chemical composition and synthesis procedure.