<h3>Purpose/Objective(s)</h3> The primary objective of this study was to identify factors associated with unexpected radiation therapy (RT) re-planning in Stage III non-small cell lung cancer (NSCLC). These factors may help predict which patients are more likely to require adaptive radical radiotherapy. <h3>Materials/Methods</h3> Stage III NSCLC patients treated in a single institution with radical intent RT from January 1, 2016, to December 31, 2019 were analyzed. Descriptive statistics were performed, including the frequency of RT re-planning and reason for re-planning. Logistic regression analysis was used to identify predictive factors associated with re-planning. Variables significant on univariate modelling, with a P value < 0.05, were selected for multivariate modelling. <h3>Results</h3> There were 144 patients with Stage III NSCLC who met study criteria. Eighteen percent (n=26) of these patients required re-planning. The most common reason for re-planning was due to volume changes (target shift or enlargement) on cone beam computed tomography (CBCT) (n=20, 77%), followed by failure to meet planning constraints (n=6, 23%). On univariate analysis, patients with a larger superior -inferior (SI) dimension of the primary and nodal planning target volume (PTV) was associated with a higher incidence of re-planning [Odds ratio (OR) 1.17, 95% CI 1.03-1.35 p = 0.02). Larger PTV (primary and nodal) was also associated with higher incidence of re-planning on univariate analysis [(OR) 2.48, 95% CI 1.21-5.38, p=0.02). On multivariable analysis, only larger PTV (primary and nodal) was statistically predictive of re-planning. <h3>Conclusion</h3> A larger SI dimension of the PTV, as well as larger PTV are associated with a higher odds ratio of re-planning. Target volume changes (tumor shift or enlargement) were the most common reasons for unanticipated re-planning. These PTV characteristics predict those stage III lung cancer targets most likely to require adaptive radiotherapy.