The idea that simplicity of explanation is important in science is as old as science itself. However, scientists often assume that parsimonious theories, hypothesis and models are more plausible than complex ones, forgetting that there is no empirical evidence to connect parsimony with credibility. The justification for the parsimony principle is strongly dependent on philosophical and statistical inference. Parsimony may have a true epistemic value in the evaluation of correlative and predictive models, as simpler models are less prone to overfitting. However, when natural mechanisms are explicitly modelled to represent the causes of biological phenomena, the application of the parsimony principle to judge the plausibility of mechanistic models would entail an unsupported belief that nature is simple. Here, we discuss the challenges we face in justifying, measuring, and assessing the trade‐off between simplicity and complexity in ecological and evolutionary studies. We conclude that invoking the parsimony principle in ecology and evolution is particularly important in model‐building programs in which models are viewed primarily as an operational tool to make predictions (an instrumentalist view) and in which data play a prominent role in deciding the structure of the model. However, theoretical advances in ecology and evolutionary biology may be derailed by the use of the parsimony principle to judge explanatory mechanistic models that are designed to understand complex natural phenomena. We advocate a parsimonious use of the parsimony principle.