How a Cornell page is laid out
The Cornell system comes from Walter Pauk, who ran the reading and study center at Cornell University and described the method in his book How to Study in College, first published in 1962 and revised through many editions since. The page is divided into three parts, and each one has a job.
The wide column on the right is for notes taken during the lecture: the ideas, examples and diagrams, written as briefly as you can manage. The narrow column on the left, the cue column, stays empty in class. Afterwards you fill it with short questions and key words that point at the notes beside them. The strip across the bottom is for a summary: two or three sentences, in your own words, on what the whole page was about.
On letter paper the classic proportions are a 2½-inch cue column and a 2-inch summary, which is what the Standard settings print. They aren't sacred. If your subject runs on definitions and formulas, a wider cue column gives your questions more room; if you summarize in a few lines, a short summary area leaves more space for notes.
Using it: Pauk's five Rs
Pauk summed up the routine as five Rs. Record during the lecture: get the main ideas and supporting facts down in the notes column, without trying to transcribe. Reduce soon afterwards, ideally the same day: read the notes and write questions or cue words in the left column, one beside each idea. Turning a paragraph into a question forces you to decide what actually matters in it.
Recite is the step most people skip, and probably the one doing most of the work. Cover the notes column, read each cue, and answer it out loud or on scrap paper before you look. Then reflect: ask how the ideas connect to each other and to what you already know, and write the summary at the bottom. Finally, review: go back to the cues every few days and recite again, for a few minutes at a time, rather than re-reading the whole page the night before the exam.
If you only have ten minutes after a lecture, spend them on the cue column. Questions written while the lecture is fresh are better than ones written a week later from notes you no longer fully understand. There's more on taking notes you'll actually use in lecture notes that work.
What the research does and doesn't show
It would be easy to oversell this. Studies that compare the Cornell format directly with other note formats are fewer and smaller than its popularity suggests, and their results are mixed. The sheet itself doesn't make you remember anything.
What is well established is the habit the sheet is built around. Pulling an answer out of memory does more for it than looking at the answer again. When Roediger and Karpicke had one group recall a prose passage and another restudy it, the recall group came out well ahead on a test a week later, even though restudying had felt more productive at the time. And reviews spread over days beat the same time spent in one sitting, a pattern Cepeda and colleagues confirmed across hundreds of experiments in their 2006 analysis. The cue column is a ready-made set of recall questions, and the review step spaces them out. That's the case for the method, and it's also why a Cornell page you never cover and recite from does little more than ordinary notes.
If you print the template, you'll be writing by hand. A widely cited 2014 study by Mueller and Oppenheimer found an advantage for longhand notes on conceptual questions, but a larger 2019 replication by Morehead, Dunlosky and Rawson didn't find a reliable difference. Write in whichever way lets you keep up, and put the effort into the recall afterwards. Why re-reading fails goes further into that.