Background: Recent data suggests that a significant number of good performance, unresectable stage III non-small cell lung cancer (NSCLC) patients do not receive standard-of-care treatment, i.e. concurrent chemoradiotherapy (cCRT) followed by durvalumab, despite being eligible. However, little is known about actionable policy barriers to delivery of cCRT to this patient population. Methods: The National Cancer Database (2004-2016) was used to identify unresected stage III NSCLC patients aged 18-79 years with Charlson comorbidity score ≤ 1. cCRT was defined as the initiations of chemotherapy (CT) and radiation therapy (RT) that were ≤14 days (n = 53,444) apart. The remaining treatment groups included sequential CRT (sCRT; n = 16,666), CT only (n = 15,416), RT only (n = 11,579), and no first course treatment (n = 16,691). Multinomial logistic regressions were used to examine the likelihoods of receiving different treatment modalities, controlling for patient demographics, Charlson comorbidity score, health insurance, facility type, social deprivation index (SDI, a comprehensive socio-economic measure; higher SDI indicates lower socioeconomic status [SES]), driving time to facility, diagnosis year, and region. Results: Of the total 113,796 patients assessed (median age 66 years), most were male (55.7%), non-Hispanic white (81.7%), and with SDI score ≥50 (51.3%). 29.5% had Charlson comorbidity score = 1 while the rest had 0. In adjusted analyses (predicted margins), 47.0% patients received cCRT (sCRT: 14.6%; CT only: 13.5%; RT only: 10.2%; no treatment: 14.7%). Compared to the privately insured, Medicaid, Medicare, and uninsured patients were more likely to receive RT only (relative risk ratios [95%CI]: 1.93 [1.77-2.11]; 1.51 [1.41-1.61]; 1.80 [1.61-2.01], respectively) and no treatment (1.84 [1.71-1.99]; 1.54 [1.45-1.63]; 2.19 [2.01-2.40], respectively) rather than cCRT (all p < .001). Moreover, higher SDI was associated with higher likelihood of receiving RT only (highest vs lowest SDI scores: 1.42 [1.33-1.52]), or no treatment (1.46 [1.38-1.55]) rather than cCRT (all p < .001). Longer driving time was associated with higher likelihood of receiving CT only ( > 120 mins vs < 30 mins: 1.24 [1.10-1.39]), or no treatment (1.33 [1.18-1.50]) rather than cCRT (all p < .001). Conclusions: Health policies should focus on patients who are not privately insured and live in neighborhoods with low SES. Moreover, helping their transportation needs may also improve the likelihood of receiving cCRT.