e21621 Background: Palliative and end of life care is essential for lung cancer pts. We examined palliative care delivery for lung cancer pts at 2 VA medical centers. Methods: In an IRB approved protocol, we reviewed medical records of matched pts diagnosed with liver or lung cancer between 2006 and 2012 who were seen by palliative care at 2 VA medical centers (S1, S2). Veterans were compared by 1) demographics, 2) palliative interventions, and 3) outcomes: ED visit within 1 month (mo) of death, ED visit within 6 mo of death, and ICU within 30 days of death. Analyses were performed with SAS Studio Version 3.5. Results: We analyzed 69 male pts (29 at S1and 40 at S2), with a mean age of 63 and 66 years respectively. Thirty (43%) pts were Caucasian and 39 (56.5%) African American. Cancer stages were I (1%), Stage II (3%), Stage III0 (29%), Stage IV (58%), and unknown (8%). Median Karnofsky performance status was 60%. By site, differences were seen in DNR/DNI (S1, S2) [ 21 (72%), 12 (30%) p = 0.0007). For palliative care interventions, significant differences were seen in treatments for pain (S1, S2) [17 (59%), 37 (93%) p = 0.0011], constipation (S1, S2) [10 (34%), 35 (88%) p < 0.0001], and dyspnea (S1, S2) [8 (28%), 29 (73%) p = 0.0003], but not for holding goals of care discussions. We observed differences between sites in evaluations by palliative social work (S1, S2) [15 (51%), 39 (98%) p < 0.0001], mental health (S1, S2) [21 (88%), 10 (25%) p < 0.0001], and chaplain visits (S1, S2) [ 19 (66%), 36 (90%) p = 0.0167]. Finally with outcome, there was a difference between consultations by other medical specialties (S1, S2) [ 19 (66%), 39 (98), p = 0.0005] but not for ED visits, admissions, procedures or ICU stays at the end of life . Conclusions: There are site specific differences between VA medical centers and may reflect local practice patterns, Additional sites should be studied.