<h3>Purpose/Objective(s)</h3> Clinical and treatment factors that engender long-term gastrostomy tube (g-tube) dependence are poorly defined. We attempted to identify potential predictors. <h3>Materials/Methods</h3> All eligible patients treated on RTOG 0129, RTOG 0522, and NRG/RTOG 1016 were grouped according to treatment received: standard fractionation RT (3DSFX) + cisplatin (C), accelerated fractionation RT (3DAFX) + C, accelerated IMRT (AFX) + C, AFX + C + cetuximab (cetux) and AFX + cetux. G-tube rates were compared at baseline (pre-treatment), 6-months after radiation completion, and annually from years 1 to 9 post treatment. Time to g- tube placement was analyzed using Cox proportional hazards models for variables of interest. Additionally, a separate review was performed for patients with no g-tube at 3 years' post-treatment (survivors with acceptable post-treatment swallowing) to evaluate late g-tube placement. <h3>Results</h3> 2389 patients were identified, 1839 of which had oropharynx cancer (OPC). AFX+C was the most common management (n=843, 35%). The G-tube rate at treatment initiation was 14%. At one-year post-treatment the rate was 19% and at 9 years the number was 8%. The RTOG 0129 g-tube rate was significantly higher than that of later studies (RTOG 0522 HR 0.7, 0.6-0.8; NRG/RTOG1016 HR 0.85 (0.72-0.99). Patients treated with 3D regimens had significantly higher feeding tube rate at baseline, 6 months, 1 year, and 2 years post-treatment (3DSFX + C: 25%/45%/29%/13% and 3DAFX + C: 22%/41%/26%/17%, respectively) when compared to accelerated IMRT regimens (all time points p < 0.001). There was no difference in feeding tube rate between treatment groups from years 3 – 9. Among OPC patients the predictors of feeding tube placement were older age (HR 1.01, 1.00-1.02), T4 primary (1.3, 1.1-1.6) and >10 pack year smoking history (HR 1.2, 1.0-1.4). Treatment regimen, tumor location, and p16-status did not affect g-tube placement. Among all patients, predictors of g-tube placement were older age (HR 1.01, 1.01-1.02); T4 primary (HR 1.3, 1.2-1.5), AJCC 7<sup>th</sup> ed N2c/N3 (1.3, 1.1-1.6); and >10 pack year smoking history (HR 1.2, 1.0-1.3). At 3 years there were 989 patients alive with follow-up who never required a gastrostomy tube. Of the 249 of these patients with 9 years' follow-up the g-tube rate was 6%. From univariable model predictors of feeding tubes placed more than 5 years after treatment completion were T3 category, T4 category, >10 pack years smoking, non-tonsil or tongue base primary, and p16-negative tumors. <h3>Conclusion</h3> G-tube rates were highest in patients managed with 3DCRT on RTOG 0129; utilization has significantly decreased over time. Radiation acceleration treatment intensification does not increase g-tube use during treatment or survivorship. Predictors of feeding tube use are increasing age, T4 primary tumors and >10 pack year smoking history.