With the gradual shift to online education models that has taken place in recent decades, research has sought to understand the nuances of student performance in an online model in comparison to more traditional in-person modalities. However, the effects of instructional modality have been difficult to determine given the many variables that exist in course design between these methods. In this study, we attempt to determine the efficacy of asynchronous online instruction by comparing two nearly equivalent courses. The first course was a flipped classroom, a recent and well-studied hybrid model of instruction. The second was an asynchronous fully online course that contained all the same instructional elements as the in-person course but lacked any student or instructor interaction. Student performance was tracked at both a highly-selective private institution and an open-enrollment public institution. Results show that students' performance drops in an asynchronous online course compared to an equivalent in-person experience. Several potential hypotheses are put forth to explain a change in performance that can potentially shape the design of online instruction.