This study aims to investigate how working memory (WM) performances and instructional strategy choices affect learners' complex cognitive task performance in online environments. Three different e‐learning environments were designed based on Merrill's (2006a) model of instructional strategies. The lack of experimental research on his framework is the main argument of this paper. The participants' WM processes with the n‐back task scores were used for defining their WM performances. This study is designed as repeated measures. Thirty‐five undergraduate students completed complex cognitive tasks three times. According to their WM groups, the participants were assigned to experimental conditions randomly by counterbalancing. The main results of the study indicated that although no performance differences were observed in complex cognitive tasks across instructional strategies, there was a statistically significant change observed across WM groups in favour of those who had high WM performances. The interaction effect did not have an effect on participants' overall performance. These results indicate that cognitive differences lead to different outcomes when the instructional design is set to be the same for all. Therefore, it is concluded that instructional design choices could take individual cognitive differences into account when developing adaptive e‐learning environments. Lay Description What is already known about this topic: E‐learning environments are often considered to be complex environments mainly because of their navigational nature (i.e., design of hyperlinks and the tools embedded within the system for personal and collaborative use). Yet, from the learners' perspective, it is not clear whether it is the content or merely the design choice that could make the online environment (more) complex. Moreover, cognitive differences play a crucial role during individuals' interactions with(in) the course materials. When measuring the performance on cognitive tasks, working memory (WM) performance is a strong predictor. What this paper adds: The literature on Merrill's instructional strategies which is one of the instructional models in designing how to teach complex tasks is mainly limited to conceptual discussions. Using different instructional strategies did not significantly change individuals' performances when dealing with complex cognitive tasks. Individuals with high level working memory performance performed significantly better. Implications for practice and/or policy: Cognitive differences lead to different outcomes when the instructional design is set to be the same for all. Studies in the field of mind, brain and education (MBE) may provide additional necessary tools not only for practitioners but also for instructional designers to better design personalized e‐learning environments in the future.