This paper presents a cognitive model of stimulus-response compatibility (SRC) effects for a situation in which location-relevant and location-irrelevant tasks are intermixed within a single trial block. We provide a computational explanation of the cognitive processing involved in the mixed-task condition. The model is based on the Instance-Based Learning Theory, developed originally to explain decision making in dynamic tasks, and the ACT-R theory of cognition. The comparison of the model's outputs to human data demonstrates high similarity, and the model offers an explanation for sequential modulations of the SRC/Simon effects observed when compatible and incompatible trials repeat or switch. Several possibilities to apply this model to novel tasks are discussed.