A key insight of associative learning theory is that learning depends on the actions of prediction error: a discrepancy between the actual and expected outcomes of a conditioning trial. This view of learning has inspired, and in turn been supported by, work in the neurosciences ranging from single unit recording and neuroimaging studies to pharmacological, chemogenetic, and optogenetic interventions. Here we review evidence describing how error-correcting learning rules are instantiated in the activity of distributed neural circuits controlling the effectiveness of unconditioned stimuli during Pavlovian fear conditioning. We show that these prediction error signals, controlling variations in event processing, are fundamental to Pavlovian fear association formation. We also argue that variations in event processing are embedded within multiplexed learning signals and that a coherent understanding of the nature and relationships between these multiple signals at specific times during the conditioning trial is needed. (PsycINFO Database Record (c) 2018 APA, all rights reserved).