The clinical utility of EUS has expanded after introducing EUS guided fine needle aspiration biopsy (FNAB). The optimal number of tissue passes needed for diagnostic certainty is unclear. While the presence of the cytopathologist (Cpt) during the FNAB is preferred, this may not be possible. The optimal number of passes needed to reach a diagnosis (dx) may provide important information to the endoscopist. This study was undertaken to 1.) determine the optimal number of passes needed during EUSFNAB to reach the correct diagnosis by evaluating the sequential sensitivity (Se), specificity (Sp), yield and certainty of the EUS-FNAB; 2.) and stratify the above by biopsy site. METHODS: 104 consecutive pts. had EUS FNAB (22 G needle) with sufficient follow-up on 96 pts. Seven or greater (=) passes/lesion were performed. The slides were read sequentially by the study Cpt who recorded the adequacy and certainty of the FNAB after each pass (Certainty score: 3- certain). The results of surgical pathology and clinical F/U served as the gold standard. The above parameters were calculated for pancreatic lesions , lymph nodes (LNs) and miscellaneous. RESULTS: Among 96 pts.,(50M/46W/mean age-62/10 month F/U) there were no complications. The graph illustrates the sequential increase in Se in pts. with a certainty score of 3. Our data shows that for pancreatic (n=34) and miscellaneous lesions, 7 passes is the optimal number of passes required to achieve a high certainty in dx. For ≥7 passes of the pancreas: Se 83.3%, Sp 100%, PPV 100%, and NPV 44.4%. For LNs (n=43) 5 passes is the optimal number to achieve an accurate dx. The operating characteristics for LN are: Se 77.3%, Sp 100%, PPV 100%, and NPV 81%. CONCLUSIONS: This study suggests that 7 FNA passes should be performed on any pancreatic lesion and 5 passes should be performed on any lymph node to ensure a high certainty of reaching a cytologic dx.