Objective: Cytopathologists’ usage patterns for ‘atypia of undetermined significance’ (AUS) in thyroid fine-needle aspiration (FNA) are not well understood. AUS rates over a 5-year period were analyzed to quantify variability and identify correlations with experience and histologic outcomes. Study Design: A retrospective review of thyroid FNAs from a tertiary-care hospital from 2005 to 2009 was performed. Results were compiled for individual cytopathologists, stratified by year, and correlated with histologic outcomes. Results: Thyroid FNAs (5,327) were evaluated by 7 cytopathologists, with an overall AUS rate of 11.2%. The annual AUS rate remained relatively constant over this time period, though notable inter- and intrapathologist variability was seen. The AUS rate was significantly lower for those with cytopathology boards (10.3%) compared to those without (14.0%). There was no correlation between the AUS rate and cytopathologist experience or thyroid FNA volume. The AUS rate and malignant outcome were inversely related: the higher an individual’s AUS rate was, the lower the rate of malignancy for that AUS cohort was. Conclusions: Individual cytopathologist AUS rates were variable and often exceeded the recommended target of 7%. The application of recently published defined diagnostic criteria, along with directed cytopathologist feedback, may reduce observer variability and appropriately lower AUS utilization.