237 Background: A number of important advancements in the treatment of metastatic non-small cell lung cancer (mNSCLC) have increased and diversified options for improved patient care based on individual characteristics. The ability to tailor therapy increases the challenges related to appropriate treatment sequencing. This study was designed to describe these emerging treatment patterns following the approval of novel targeted agents. Methods: Flatiron Health OncoEMR, a nationally-representative electronic medical records database in the US, was used to evaluate treatment patterns by histology (squamous and nonsquamous). Eligible patients were 18+ years of age who initiated 2ndline therapy for mNSCLC from Dec 2014-Jul 2016. Descriptive statistics were used to characterize the clinical and demographic characteristics of the study population, treatments used by line of therapy, and treatment sequencing. Analyses were conducted using SAS version 9.2. Results: A total of 3498 eligible patients were included in this study: 51.3% male; mean age 66.6 years; 65% white; 25% squamous/70.7% nonsquamous (4.3% not specified); and 93% were treated at community practices. ALK testing was performed on 20.0%/74.8%, EGFR testing on 21.5%/79.8%, and PDL-1 on 8.6%/9.7% of patients with squamous/nonsquamous tumors, respectively. Single-agent PDL-1 inhibitors were used by 54.2% of squamous and 35.2% of nonsquamous patients in the 2nd-line setting; however, there were more than 35 (squamous) and 64 (nonsquamous) unique first-line regimens prior to single-agent PDL-1 treatment. Other 2nd-line regimens included pemetrexed (24.9% of nonsquamous patients) and gemcitabine (18.4% of squamous patients), which were preceded by 70 and 48 unique first-line regimens, respectively. Conclusions: There is interest in understanding treatment sequencing to identify the optimal sequence of care for patients with NSCLC; however, there was considerable heterogeneity in sequencing. Since few patients follow any similar trajectory of care, comparative effectiveness research will be challenged to appropriately balance groups due to insufficient patient numbers in any specific treatment sequence.