Although a general understanding of many aspects of fatigue crack growth behavior was established in the early 1960s, a specific 'accumulation of damage model' for computation of growth under a wide variety of service loads was lacking. The control of growth rates by K, the crack tip stress intensity factor and its reversing plastic zone, was well understood but somehow crack closure was overlooked until W. Elber's astute observations in the late 1960s. Various 'service load damage models' were proposed in the 1970s which at best successfully augmented extensive testing under simulated service load conditions. More recently, the crack closure/finite element model of J. Newman has shown the greatest promise. However, the precise cyclic load crack growth data of J.K. Donald has shown limitations of the 'crack opening based ⌬K' as the most appropriate damage criterion. A simple partial crack closure model will be explored herein to attempt to better understand the near-threshold effects observed. Although this new model, as an augmentation of Newman's model, shows great promise, both of these models require measurements of opening loads which are avoided by Donald's 'adjusted compliance ratio' and other approaches. Further possibilities for modeling will be discussed in this historical context to attempt to anticipate future developments. Certainly, some simplification through physical understanding is still required. The recent development of a simplified view of cyclic crack growth behavior by R.W. Hertzberg will be cited as an example of ultimate relevant simplification. Optimism that better understanding will lead to yet better and simpler 'accumulation of service load damage models' will be emphasized.