A review is presented of the history of the Point Defect Model (PDM) for the growth and breakdown of passive films that form on the surfaces of reactive metals in contact with corrosive, condensed phase environments. The PDM has passed through three generations, with each successive generation addressing issues that have arisen from experiment or theory. Thus, the first generation model (PDM-I), developed in the early 1980s, assumed that the passive film was a single defective oxide layer that contained cation vacancies and oxygen vacancies that were generated and annihilated at the metal/film and film/solution interfaces, as inspired by the work by Wagner on high temperature oxidation. As with gas-phase systems, the film was assumed not to dissolve. However, it soon became evident that this model could not account for the properties of the passive state on metals in contact with aqueous environments and, accordingly a Generation II model (PDM-II) was developed to address these issues. PDM-II incorporated the bi-layer structure of the film comprising a defective oxide or hydride barrier layer adjacent to the metal and an outer layer that forms by precipitation of material from reaction of transmitted cations with species in the environment (including water, CO 3 2-, HS -, etc.), introduced metal interstitials to the suite of defects, recognized barrier layer dissolution, and recognized the need to classify reactions as to whether they are lattice conservative or non-conservative, but assumed that control of the passive current resided in the barrier layer alone. PDM-II has enjoyed considerable success and the author knows of no instance where it has been demonstrated to be at odds with experiment when confluence between experiment and theory has been demonstrated. A Generation III model (PDM-III) has been recently developed to extend the theory to those cases (e.g., the valve metals) where the outer layer is so resistive that it controls the impedance of the interface and hence the corrosion rate. A fourth generation model that will describe passivity on alloys is now under development. The experimental evidence upon which each generation is based is reviewed.