Learning in Massive Open Online Courses (MOOCs) requires learners to self-regulate their learning process or receive effective self-regulated learning (SRL) interventions to accomplish personal goals. Much attention has thus been paid to how SRL influences learner performance in MOOCs, but research has overlooked a person-centered analysis of how online learners perform SRL in this setting. Without understanding this individual difference, educators are unlikely to provide effective SRL interventions tailored to each type of self-regulated learner. In addition, it remains uncertain how culture shapes the differences in SRL traces, especially given that most of the existing understandings of SRL are rooted in Western cultures. To fill the gaps, this research applied learning analytics to explore learner profiles in terms of how they performed SRL in MOOCs. Using K-means clustering analysis, this research revealed four different self-regulated learner profiles: all-around SRL learners, disillusioned SRL learners, control-oriented SRL learners, and control-dominated SRL learners. In addition, all-around SRL learners outperformed the other three clusters in course grades. This research also identified cultural differences between those clusters. Practical implications on how to design effective SRL interventions are provided.