With improvements in abdominal imaging, detection of incidental pancreatic cysts are becoming increasingly common. Analysis of pancreatic cyst fluid from fine-needle aspiration is particularly important in identifying intraductal papillary mucinous neoplasms (IPMNs) and mucinous cystic neoplasms (MCNs), which have significant implications in clinical intervention and follow-up. Previous controlled studies have shown that KRAS mutations in cyst fluid are highly specific for mucinous differentiation in pancreatic cysts; however, this has not been examined in the clinical setting. Over a 6-year study period, 618 pancreatic cyst fluids obtained by fine-needle aspiration at the time of endoscopic ultrasound were tested for KRAS mutations as part of routine evaluation for a cystic neoplasm. Of the 618 specimens, 603 (98%) from 546 patients were satisfactory for molecular analysis. Patients ranged in age from 17 to 90 years (mean, 63.9 years) and were predominantly female (68%). Pancreatic cysts were relatively evenly distributed throughout the pancreas and ranged in size from 0.6 to 11.0 cm (mean, 2.3 cm). Mutations in KRAS were detected in 232 of 603 (38%) aspirates. Although sufficient for molecular analysis, 320 of 603 (53%) specimens were either less than optimal (38%) or unsatisfactory (15%) for cytopathologic diagnosis. Surgical follow-up information was available for 142 (26%) patients and consisted of 53 KRAS-mutated and 89 KRAS-wild-type cysts. Overall, KRAS mutations had a specificity of 100%, but a sensitivity of 54% for mucinous differentiation. When stratified by cyst type, KRAS had a sensitivity of 67% and 14% for IPMNs and MCNs, respectively. In summary, KRAS mutations were highly specific for mucinous differentiation, but were inadequate in identifying MCNs. Future molecular studies and the combination of other fluid markers are required to improve the detection and classification of pancreatic mucinous neoplasms by endoscopic ultrasound fine-needle aspiration.