Rapid on-site evaluation (ROSE) of endobronchial ultrasound-guided transbronchial needle aspirates (EBUS-TBNA) has not been compared to final detailed cytological analysis in patients with suspected sarcoidosis. To assess the diagnostic accuracy of EBUS-TBNA with ROSE in patients with suspected sarcoidosis, a prospective two-centre study performed EBUS-TBNA with ROSE of cellular material followed by transbronchial lung biopsy (TBLB) and endobronchial biopsy (EBB). The diagnostic accuracy of EBUS-TBNA with ROSE was compared to the final cytological assessment and to TBLB and EBB. Analysis confirmed 49 out of 60 cases of sarcoidosis. ROSE sensitivity was 87.8% (specificity 91%, positive predictive value 97.7%). ROSE slide interpretation in combination with the final fixed slide and cell block preparations had a sensitivity of 91.8% (specificity 100%, positive predictive value 100%). 67% of patients were confirmed as having sarcoidosis on TBLB and 29% on EBB. Interobserver agreement between cytotechnologists and pathologists was very good (κ=0.91, 95% CI 0.80-1.0 and κ=0.91, 95% CI 0.79-1.0, respectively). EBUS-TBNA with ROSE has high diagnostic accuracy and interobserver agreement and informs the bronchoscopist in theatre whether additional diagnostic procedures need to be undertaken. EBUS-TBNA with ROSE should therefore be considered as the first-line investigation of sarcoidosis.