Background <p>Sarcoidosis is a heterogeneous granulomatous disease with an unpredictable course. Population-level evidence describing its long-term trajectories and their prognostic implications remains limited.</p> Methods <p>We identified individuals newly diagnosed with sarcoidosis in the Swedish National Patient Register (≥ 2 ICD-coded visits; 2006–2018). Sarcoidosis-related visits over five years were modeled using zero-inflated Poisson finite mixture models to identify trajectories. Baseline demographic and clinical predictors of trajectory membership were assessed with Poisson regression. Associations between trajectories and long-term outcomes were estimated using multivariable Cox regression. Supplementary analyses incorporated clinical, genetic, and physiological data from the Karolinska cohort.</p> Results <p>Among 9665 patients, two distinct trajectories were identified: a resolving trajectory (71.5%) with near-complete remission of sarcoidosis-related visits within two years, and a chronic trajectory (28.5%) with persistently elevated visit rates over five years. Older age modestly increased the risk of chronic disease. The strongest predictor of chronic disease was immunosuppressive treatment around diagnosis (risk ratio 2.18 [95% CI 2.04, 2.33]). Diagnosis in neurology, ophthalmology, or cardiology, uveitis, and hospitalization at diagnosis were also associated with chronicity, whereas diagnosis in rheumatology and dispensation of non-steroidal anti-inflammatory drugs were protective. In the Karolinska cohort, Löfgren’s syndrome and HLA-DRB1*03 were strongly associated to a resolving course. Chronic sarcoidosis was associated with higher risks of infection, heart failure, diabetes, depression, anxiety, and early death.</p> Conclusions <p>Sarcoidosis follows two long-term trajectories. Clinical phenotype and need of early treatment predict chronicity, which is associated with substantial long-term morbidity and mortality. Early prognostication may support personalized sarcoidosis management.</p>