This dissertation explored the influence of representational complexity on the observation of linear retention and forgetting patterns of episodic memory. Representational complexity was manipulated through the degree of original learning and the type of processing engaged in at encoding. Episodic retention is commonly thought to follow a negatively accelerating pattern, often described by a power function. However, Fisher and Radvansky (2019) showed that linear forgetting patterns can be reliably observed. Considering the results of our own experiments, as well as those from the literature, a number of factors emerge as likely contributors to the observation of linear forgetting. Two of these are (a) the degree of original learning at encoding, and (b) the amenability of the materials to representationally complex traces. This dissertation reports four experiments, with two each addressing each of these two potential contributors. The results revealed that original learning that involves either retrieval practice or elaboration can contribute to producing linear retention.