Ophir, Nass and Wagner (2009) showed that as multitasking frequency increases, multitasking performance decreases. Other studies, however, have not replicated this effect (e.g., Minear, Brasher, McCurdy, Lewis & Younggren, 2013). In this paper, we argue that the association between frequent media multitasking and poor multitasking performance depends on self-regulation ability and external factors, such as manipulation of the task execution strategy (sequential vs. free switching). In Study 1, we determined participants’ media multitasking frequency and measured their self-regulation ability. Then, participants performed a multiple media task in which they could freely switch between browser tabs. The results showed that high media multitasking levels were associated with more switches between tabs but only for participants with low (but not high) self-regulation ability. No differences in performance were observed. In Study 2, instead of measuring self-regulation ability, we manipulated task execution strategy (as an external form of regulation). As predicted, media multitasking frequency and performance on multiple tasks (overall score) were negatively related only in the free switching condition and not in the sequential condition. The results elucidate the relationship between media multitasking frequency and multitasking performance by showing its boundary conditions, and they help explain contradictory findings in the media multitasking literature.