Abstract Background The immediate assistance of a cytologist during EUS-guided FNA is not universal. The optimal number of fine needle passes during EUS-guided FNA has not been determined in a prospective study. The aim of this study was to determine the optimal number of passes required to obtain a correct diagnosis. Methods Seven or more passes were made with a fine needle into a variety of lesions during EUS-guided FNA. Adequacy of the aspirate, diagnosis, and a “certainty score” were recorded after each pass and interpreted sequentially by a cytopathologist. Surgical histopathology and 1-year clinical follow-up were used as reference standards. The percentage of correctly diagnosed cases was calculated and stratified according to organ, disease group, and EUS characteristics of the lesion. Results Lesions from 95 patients were categorized into the following locations: pancreas, lymph node, and miscellaneous. The sensitivity and specificity for 7 passes from the pancreas and miscellaneous lesion groups were, respectively, 83% and 100%. The sensitivity and specificity for 5 passes from the lymph node group were, respectively, 77% and 100%. Conclusions During EUS-guided FNA, at least 7 passes with a fine needle into pancreatic and miscellaneous lesions, and 5 passes into lymph nodes are needed to ensure a high degree of certainty for making a correct diagnosis.