Learning mathematics and science entails learning the relations among multiple interacting elements, especially when solving problems. Assimilating multiple interacting elements simultaneously in the limited working memory capacity would incur cognitive load. Unless the instructions provide a mechanism to manage the high cognitive load involved, learning effectiveness may be compromised. Researchers have investigated instructional efficiency across diverse domains from the perspective of cognitive load theory. Progress in educational theory has enabled a better understanding of three types of cognitive load that students experience during the learning process: intrinsic, extraneous, and germane cognitive load. Processing the intrinsic nature of a task constitutes intrinsic cognitive load (e.g., complexity of elements). Sub-optimal instruction requiring unnecessary processing of elements constitutes extraneous cognitive load. Investing mental effort in multiple practices constitutes germane cognitive load. Recent advance in cognitive load theory highlights element interactivity (i.e., the interaction among elements to be processed) as a common thread among different types of cognitive load. However, despite progress in cognitive load research, little is known about the effects of element interactivity in secondary school mathematics and science education. Using element interactivity as a point of reference, this article reviews the design features of different approaches to teaching linear equations in mathematics and the topic of density in science. Evidence seems to point to the practical benefit of using instructional approaches that address the issue of multiple elements interacting with each other to facilitate learning. As such, the conceptualization of cognitive load in terms of element interactivity will bring further progress in the research on cognitive load in mathematics and science learning.