Sarcoidosis is a multi-systemic granulomatous disease. Affected individuals can show spontaneous healing, develop remission with drug treatment within 2 years, or become chronically ill. Our main goal was to identify features that are related to prognosis.The study consisted of 101 patients, recruited at a single center, who were already diagnosed with sarcoidosis at the start of the study or were diagnosed within 48 months. Ninety individuals were followed-up for at least 24 months and were classified according to clinical outcome status (COS 1 to 9). Those with COS 1-4 and COS 5-9 were classified as having favorable and unfavorable outcomes, respectively. Unconditional logistic regression analyses were conducted to define which variables were associated with sarcoidosis outcomes. Subsequently, we established a scoring system to help predict the likelihood of a favorable or unfavorable outcome.Of our patients, 48% developed a chronic form of the disease (COS 5-9). Three clinical features were predictive of prognosis in sarcoidosis. We built a score-based model where the absence of rheumatological markers (1 point), normal pulmonary functions (2 points), and the presence of early respiratory symptoms manifestations (2 points) were associated with a favorable prognosis. We predicted that a patient with a score of 5 had an 86% (95% confidence interval [CI] 74%-98%) probability of having a favorable prognosis, while those with scores of 4, 3, 2, 1, and 0 had probabilities of 72% (95% CI 59-85%), 52% (95% CI 40-63%), 31% (95% CI 17-44%), 15% (95% CI 2-28%), and 7% (95% CI 0-16%) of having a favorable prognosis, respectively. Thus, our easy-to-compute algorithm can help to predict prognosis of sarcoidosis patients, facilitating their management.