Cognitive load theory has been in development since the 1980s. Much of the impetus for that development has come from firstly, replication failures using randomised controlled trials and secondly, from the incorporation of other theories into cognitive load theory. Both have led to theory expansion. The immediate cause of the so-called "replication crisis" in psychology and other disciplines is a failure to replicate previous empirical findings. Using cognitive load theory as an example, I argue that the appearance of contradictory evidence does not necessarily derive from a failure to properly collect data. Rather, it can be caused by initially insufficiently detailed theories, with increasing detail often revealing the reason for a failure to replicate. For cognitive load theory, each failure to replicate, rather than being a negative, contributed to the further development of the theory. In addition, the theory has developed over many years by closely incorporating other theories associated with human cognitive architecture and evolutionary psychology. In this paper, I discuss some of the developmental milestones associated with cognitive load theory and how they were informed by replication failures and theory integration.