Objective: Evidence for the utility of medications in settings lacking randomized trial data can come from studies of treatment persistence. The present study was undertaken to examine patterns of medication use in systemic lupus erythematosus (SLE) using data from a large multicenter longitudinal cohort. Methods: Prospectively collected data from the Asia Pacific Lupus Collaboration cohort including disease activity (SLE Disease Activity Index 2000 [SLEDAI-2K]) and medication details, captured at every visit from 2013–2018, were used. Medications were categorized as glucocorticoids (GCs), antimalarials (AM), and immunosuppressants (IS). Cox regression analyses were performed to determine the time-to-discontinuation of medications, stratified by SLE disease activity. Results: Data from 19,804 visits of 2,860 patients were analyzed. Eight medication categories were observed: no treatment; GC, AM, or IS only; GC plus AM; GC plus IS; AM plus IS; and GC plus AM plus IS (triple therapy). Triple therapy was the most frequent pattern (31.4% of visits); single agents were used in 21% of visits, and biologics in only 3%. Time-to-discontinuation analysis indicated that medication persistence varied widely, with the highest treatment persistence for AM and lowest for IS. Patients with a time-adjusted mean SLEDAI-2K score of ≥10 had lower discontinuation of GCs and higher discontinuation of IS. Conclusion: Most patients received combination treatment. GC persistence was high, while IS persistence was low. Patients with high disease activity received more medication combinations but had reduced IS persistence, consistent with limited utility. These data confirm unmet need for improved SLE treatments.