Satralizumab showed a comparable safety profile versus placebo in 2 pivotal neuromyelitis optica spectrum disorder (NMOSD) studies. We analyzed infection rates with long-term satralizumab treatment in the open-label study, SAkuraMoon, and in a post-marketing setting (PMS), comparing frequencies with US-based health claims real-world data (US-RWD). Incidence rates of infection per 100 patient-years (IR/100 PY) were analyzed in the SAkura studies (clinical cut-off date: 31 January 2023). Reported rates of infection ( %) in a PMS using Periodic Benefit-Risk Evaluation Reports (2020-2023), and cumulative incidence of infections ( %) from the US PharMetrics claims data in NMOSD patients (2017-2022) were analyzed. 166 patients (SAkura studies), 2951 patients (PMS) and 2872 patients (US-RWD) were included. In the SAkura studies, the incidence rates of infection, serious infection, and sepsis were lower versus the double-blind period (IR/100 PY [95 % confidence intervals (Tur, C. et al.)] infection 91.7 [85.5-98.3] vs 113.0 [98.6-129.0]; serious infection 2.6 [1.7-3.9] vs 4.1 [1.8-8.1]; sepsis 0.6 [0.2-1.3] vs 1.0 [0.1-3.7], respectively). In a PMS, reported rates of infection, serious infection, and sepsis were 7.3 %, 3.8 %, and 0.6 %, respectively. In the US-RWD, cumulative incidence of infection, serious infection, and sepsis in NMOSD were 67.3 %, 8.4 %, and 4.9 %, respectively. Concomitant IST use, comorbidities, Expanded Disability Status Scale score ≥4.0, and age >65 years were potential confounders of sepsis. US-RWD indicated infection is a major comorbidity in NMOSD, independent of satralizumab treatment. Infection rates were consistently lower in satralizumab-treated patients compared with US-RWD. Trial Registration: NCT04660539(SAkuraMoon), NCT02028884(SAkuraSky), NCT02073279(SAkuraStar).