We investigate the impact of a compressed, immersive teaching format on student learning in online courses based on a validated measure of learning outcomes. This study focuses on larger enrollment (50+) introductory undergraduate geology courses since 2007 that were evaluated with the Geoscience Concept Inventory (GCI). The impact of a course on student performance is estimated by calculating the normalized improvement, comparing end of course GCI scores with start of course GCI scores for each student. There were increased student performance gains that are both large and statistically significant following each of our main course revisions: from courses with a traditional lecture-based approach (9.2 ± 1.0%) to courses with an active learning revision (17.2 ± 0.8%), and then from those classroom-based active learning to courses with active learning delivered as hybrid or fully online (24.5 ± 2.0%). The improvements have been interpreted to be due to purposeful re-design of the courses to focus more on student engagement. The change to the e-learning format involved developing a wide range of new course components including: assignments each class period with automatic grading and feedback, regular use of software (Google Earth and Microsoft Excel) with video tutorials, and more realistic scientific investigations using authentic internet databases. We are able to isolate the impact of immersion in these courses, as fully online versions of the course were offered in both a traditional 15-week semester and a 3-week winter term, with essentially no change in course design other than the time frame. The GCI results indicate that student performance in the compressed 3 week course (30.3 ± 3.6%) was considerably higher than that in the 15 week course (19.7 ± 3.6%). These results indicate institutions should consider focusing e-learning efforts outside of the traditional 15-week semesters when compressed, immersive courses can be offered more readily.