Depletion of limited working memory resources may occur following extensive mental effort resulting in decreased performance compared to conditions requiring less extensive mental effort. This "depletion effect" can be incorporated into cognitive load theory that is concerned with using the properties of human cognitive architecture, especially working memory, when designing instruction. Two experiments were carried out on the spacing effect that occurs when learning that is spaced by temporal gaps between learning episodes is superior to identical, massed learning with no gaps between learning episodes. Using primary school students learning mathematics, it was found that students obtained lower scores on a working memory capacity test (Experiments 1 and 2) and higher ratings of cognitive load (Experiment 2) after massed than after spaced practice. The reduction in working memory capacity may be attributed to working memory resource depletion following the relatively prolonged mental effort associated with massed compared to spaced practice. An expansion of cognitive load theory to incorporate working memory resource depletion along with instructional design implications, including the spacing effect, is discussed.