We investigate the role of metal-ligand bond fission in the nonradiative decay of excited states in iridium(III) complexes with applications in blue organic light-emitting diodes (OLEDs). We report density functional theory (DFT) calculations of the potential energy surfaces upon lengthening an iridium-nitrogen (Ir-N) bond. In all cases we find that for bond lengths comparable to those of the ground state the lowest energy state is a triplet with significant metal-to-ligand change transfer character ((3)MLCT). But, as the Ir-N bond is lengthened there is a sudden transition to a regime where the lowest excited state is a triplet with significant metal centered character ((3)MC). Time-dependent DFT relativistic calculations including spin-orbit coupling perturbatively show that the radiative decay rate from the (3)MC state is orders of magnitude slower than that from the (3)MLCT state. The calculated barrier height between the (3)MLCT and (3)MC regimes is clearly correlated with previously measured nonradiative decay rates, suggesting that thermal population of the (3)MC state is the dominant nonradiative decay process at ambient temperature. In particular, fluorination both drives the emission of these complexes to a deeper blue color and lowers the (3)MLCT-(3)MC barrier. If the Ir-N bond is shortened in the (3)MC state another N atom is pushed away from the Ir, resulting in the breaking of this bond, suggesting that once the Ir-N bond breaks the damage to the complex is permanent-this will have important implications for the lifetimes of devices using this type of complex as the active material. The consequences of these results for the design of more efficient blue phosphors for OLED applications are discussed.