Angiographic vasoconstriction in reversible cerebral vasoconstriction syndrome (RCVS) is often undetectable at symptom onset and the diagnosis relies on clinical presentation. Although thunderclap headache is a hallmark feature of RCVS, the incidence and predictors of long-term headaches (LTH) are incompletely understood. Our study aims were twofold: to examine the sensitivity and specificity of a recently developed score (RCVS2) for vasoconstriction detection in a real-world clinical context and describe the incidence and predictors of LTH beyond the acute phase of RCVS. Retrospective analysis of consecutive patients with clinical diagnosis of RCVS in a tertiary hospital between 2017 and 2021. We examined associations between demographic factors, comorbidities, medications, imaging characteristics, and LTH (defined as at least one episode present at greater than 6-months follow-up necessitating medication). We separately examined the association between RCVS2 score and angiographic vasoconstriction and computed its sensitivity, specificity, and negative and positive predictive value based on established cutoffs (certain ≥ 5, negative ≤ 2). We included 55 patients, 50.5 (± 13.7) years; 41 (75%) female. 25 (49%) patients had LTH; only prior history of headache was significantly associated with LTH [OR 4.3, 95% CI (1.1-16.2), p = 0.03]. We found a significant association between RCVS2 score and angiographic vasoconstriction [OR 1.49, 95% CI (1.18-1.88), p = 0.001]; sensitivity, specificity, and positive and negative predictive value were 64%, 94%, 95% and 58% respectively. Approximately 50% of RCVS patients experienced LTH; only prior headache history was associated with its incidence. The RCVS2 score had a significant association with high specificity and positive predictive value for angiographic vasoconstriction in our cohort, validating its utility in improving the accuracy of diagnosis in the clinical setting.