PURPOSE: To evaluate the diagnosis improvement of EUS-FNA when using ROSE performed by the endosonographer. METHODS: A retrospective study was conducted. A total of 48 pancreatic solid masses EUS-FNA were divided into two groups according to the availability of on-site cytology (ROSE) - the first 24 patients (group A-without ROSE) and the latter 24 cases (group B-with ROSE). Sensitivity, specificity, positive predictive value, negative predictive value, accuracy, complications and inadequacy rate of EUS-FNA were determined and compared. RESULTS: Among the 48 EUS-FNA, the overall performance was: sensitivity 82%; specificity 100%; positive predictive value (PPV) 100%; negative predictive value (NPV) 70% and accuracy 87%. The sensitivity of the Group A was 71%, versus 94% in-group B (p=0.61). Moreover, the negative predictive value was 58% versus 87% (p=0.72). The accuracy rate increased from 79% to 96% (p=0.67) in the ROSE group. The number of punctures was similar between the groups. No major complications were reported. CONCLUSION: Rapid on-site cytopathological examination, even when performed by the endosonographer, may improve the diagnostic performance in the diagnosis of solid pancreatic lesions, regardless of the slight increase in the number of punctures.