Objectives It is unknown if hydroxychloroquine blood level dynamics impact flare rates in lupus nephritis patients. We prospectively evaluated hydroxychloroquine levels to determine which blood-based patterns are more associated with disease activity. Methods In total, 82 lupus nephritis patients under a prescribed hydroxychloroquine dose of 4–5.5 mg/kg actual body weight (maximum 400 mg/day) for ≥3 months were evaluated at baseline and 7 months. Hydroxychloroquine blood levels were determined by liquid chromatography-tandem mass spectrometry. Flare was defined as increase ≥3 in the Systemic Lupus Erythematosus Disease Activity Index 2000 score and/or a change or increase in therapy. Results Overall, 9/82(11%) patients had flares during follow-up and had lower baseline hydroxychloroquine blood levels than those without flares (220.4 (53.5–1471.1) vs. 1006.3 (53.5–2137.8) ng/ml, p = 0.013). The hydroxychloroquine blood level cut-off that best predicted flares was 613.5 ng/ml (odds ratio = 8.67, 95% confidence interval: 1.66–45.18, p = 0.006). For 77 (94%) patients, the 7-month hydroxychloroquine level dynamics was evaluated and revealed: 59/77 (77%) had a persistent pattern of adequate (41/77(53%)) or fluctuating (18/77 (23%)) levels, with a low and comparable risk of flares (2/41 (5%) vs. 1/18 (5%), p = 1.000). The remaining group had persistent low levels (18/77 (23%)) and their flare frequency was significantly higher than the adequate group (5/18 (28%) vs. 2/41 (5%), p = 0.023). The frequencies of adequate/inadequate hydroxychloroquine blood levels in patients were comparable at baseline and 7 months (McNemar’s test, p = 0.480). Conclusion We provide novel evidence that hydroxychloroquine blood-level patterns (persistently low, adequate, or intermittent) have distinct impacts on flare rates in lupus nephritis patients. These findings reinforce the need of routine hydroxychloroquine measurements to maintain the appropriate blood levels.