Objectives We sought to further determine whether cognitive test results changed for advanced cardiac life support (ACLS) taught in the team-based learning/flipped classroom format (TBL/FC) versus a lecture-based (LB) control. Methods We delivered 2010 ACLS to two classes of fourth-year medical students in the TBL/FC format (2015-2016), compared to three classes in the LB format (2012-2014). There were 27.5 hours of instruction for the TBL/FC model (TBL - 10.5 hours, podcasts - nine hours, small-group simulation - eight hours), and 20 hours (lectures - 12 hours, simulation - eight hours) in LB. We taught TBL for 13 cardiac cases while LB had none. Didactic content and seven simulated cases were the same in lecture (2012-2014) or in podcast formats (2015-2016). Testing was the same using 50 multiple-choice (MC) format questions, 20 rhythm-matching questions, and seven fill-in management of simulated cases. Results Some 468 students enrolled in the course 259 (55.4%) in the LB format in 2012-2014, and 209 (44.6%) in the TBL/FC format in 2015-2016. The scores for two out of three tests (MC and fill-in) increased with TBL/FC. Combined, median scores increased from 93.5% (IQR 90.6, 95.4) to 95.1% (92.5, 96.8, p = 0.0001). More students did not pass one of three tests with LB versus TBL/FC (24.7% versus 18.2%), and two or three parts of the test (8.1% versus 4.3%, p = 0.01). On the contrary, 77.5% passed all three with TBL/FC versus 67.2% with LB (change 10.3%, 95% CI 2.2%-18.2%). Conclusion TBL/FC teaching for ACLS improved written test results compared with the LB format.