Background: Recent trials have demonstrated improved outcomes in the 1st-line treatment of advanced pancreatic cancer (APC). However, there is limited randomized data to guide 2nd-line chemotherapy (CT) selection. We aimed to characterize predictors and outcomes of 2nd-line CT in patients (pts) with APC. Methods: We identified all pts with APC (locally advanced (LAPC) or metastatic (MPC)) who received ≥1 cycle of 1st-line CT between January 1, 2012 and December 31, 2015 across 6 centers in British Columbia, Canada. Baseline characteristics and survival outcomes were summarized. Results: Of 676 pts with APC (31% LAPC, 69% MPC) who received ≥1 cycle of CT, 164 (24%) received 2nd-line CT. These pts were younger (median 63.7 vs. 67.4 years; p= 0.01), had a lower ECOG (77% ECOG 0-1 vs. 51% ECOG ≥2; p< 0.001), and higher CA19-9 (median 1034 vs. 829; p= 0.01) compared to patients who did not receive 2nd-line CT. There were no differences in rates of 2nd-line CT between LAPC and MPC (28% vs. 23%; p= 0.18). On logistic regression, only 1st-line FOLFIRINOX (OR 5.90, p< 0.001) was associated with 2^nd line CT. CT regimens are summarized by line (Table). Median duration of 2nd-line CT was 3 cycles (range 1-30). Median overall survival (mOS) from diagnosis of patients with 2^nd-line CT was 16 months. mOS from 2nd-line CT was longer with 2nd-line gemcitabine/nab-paclitaxel than fluoropyrimidine or gemcitabine (7.9 vs. 5.1 vs. 4.3 months; p= 0.008). On multivariate analysis, longer OS from 2nd-line CT was associated with gemcitabine/nab-paclitaxel (vs. single agent CT), lower ECOG, LAPC (vs MPC), and lower CA 19-9 (HRs 0.49, 0.67, 0.58, 0.38, respectively). Conclusions: In this population-based cohort, pts treated with 2nd line CT were younger, have better ECOG, similar rates of LAPC vs. MPC, and achieved a median OS of 16 months. 1st-line FOLFIRINOX was the strongest predictor of 2nd-line CT. Gemcitabine/nab-paclitaxel was associated with superior 2nd line OS compared to gemcitabine/fluoropyrimidine.