Biopsies carry significant risk of hemorrhage, pneumothorax and infection. Using the Surveillance, Epidemiology, and End Results (SEER) – Medicare database, we hypothesize that pre-treatment diagnostic biopsy may increase morbidity rates without affecting mortality rates in patients with stage I NSCLC treated with definitive external beam radiotherapy (EBRT), especially in more frail, oxygen (O2) dependent patients. Patients in our cohort were (1) 365 years old; (2) diagnosed with AJCC6 stage I NSCLC between 2002 and 2011; (3) covered by Part A and B one year before to one year after diagnosis without supplemental HMO coverage; (4) not in a nursing home at diagnosis; and (5) treated with definitive radiotherapy without surgery. Chi-square testing compared toxicity rates from date of diagnosis or biopsy to 6m after that date in (1) patients regardless of O2 dependency status that did and did not receive diagnostic biopsy, and (2) any biopsied patient with and without O2 dependency at diagnosis. Toxicities and O2 dependency were determined using Medicare claims and ICD-9 diagnosis codes (see Table). OS and CSS were compared using the Kaplan-Meier and log rank tests. 3,231 patients met criteria in the Methods, above. Of these, 68 (2%) were diagnosed clinically without confirmatory biopsy. Patients without a prior biopsy were more likely to be white (97 v 87%, p = 0.02) and have COPD (78 v 56%, p < 0.01). The groups did not differ in Charlson score (18 v 24% with Charlson > 2, p = 0.21); marital status (51 v 49%, p = 0.77); O2 dependency (60 v 54%, p = 0.27); socioeconomic status (p = 0.17); being homebound (4% v 4%, p = 0.83); lobe involved (p = 0.38); sex (53 v 50%, p = 0.67); or median age (77 v 76y, p = 0.22). Biopsied and O2 dependent patients suffered more toxicity (see Table). In unadjusted analysis, there was no difference in median OS (23 v 21m, p = 0.53) or CSS (30 v 30m, p = 0.35) in EBRT patients between the non-biopsy and biopsy arms.Abstract 2998; TableNo Biopsy n=68Biopsy n=3,163pNo O2 n=1,464O2 n=1,699pHemorrhage0 (0%)92 (3%)0.1533 (2%)59 (4%)0.04Intubation1 (1%)126 (4%)0.2933 (2%)94 (6%)<0.01MI-->0.2561 (4%)100 (6%)0.03Pneumothorax8 (12%)935 (30%)<0.01285 (19%)473 (28%)<0.01Pneumonia-->0.2519 (1%)64 (4%)<0.01Pulm Embolus1 (1%)41 (1%)0.911 (1%)31 (2%)0.01Pulm Insufficiency0 (0%)78 (3%)0.1925 (2%)53 (3%)0.01Sepsis2 (3%)229 (7%)0.1776 (5%)156 (9%)<0.01Ventilator1 (1%)138 (4%)0.2440 (3%)99 (6%)<0.01 Open table in a new tab We found increased morbidity in biopsied patients – especially O2 dependent ones – with similar mortality compared to unbiopsied or non-O2 dependent patients. Future studies should prospectively evaluate the efficacy of omitting diagnostic biopsies in select patient populations, particularly patients with multiple co-morbidities or baseline respiratory compromise.