If you have ever finished a study day and felt like the material leaked out overnight, the problem is almost never your memory. It is the technique. Learning science has compared the popular study methods head to head for decades, and the results are consistent and a little uncomfortable: the techniques that feel best tend to work worst. Here is the honest ranking, applied to bar prep.
Producing material from memory, rather than looking at it again, is the most consistently supported finding in the learning literature. For bar prep it has a bonus property: it is the exam. Essays ask you to produce rules on a blank page, so every retrieval rep is also a dress rehearsal. The full write, compare, and grade loop is in our guide on how to memorize bar exam rules.
Reviews spread over days beat the same reviews massed into one afternoon, because each well-timed retrieval flattens the forgetting curve. Spacing decides when you study; retrieval decides how. Together they are the backbone, and everything below is decoration. Intervals that fit an eight to ten week prep are in our spaced repetition guide.
Working memory holds only a few items, so group raw material into meaningful units. In bar prep, the natural chunk is the element. A rule like negligence is not forty words; it is four slots: duty, breach, causation, damages. Store rules as element checklists and you have chunked them correctly by default. This is exactly why our deck breaks every rule into element-level rule statements.
Acronyms like MY LEGS and OCEAN are retrieval handles for list-shaped rules. They recall the checklist; the rule statements behind each letter earn the points. Use the classics, build your own sparingly, and never confuse knowing the acronym with knowing the rule. The full treatment is in our bar exam mnemonics guide.
Close the outline and write everything you can remember about a topic on a blank page, then compare against the source. It is retrieval practice in bulk: cheap, brutal, and honest. One blurt per subject per week is an excellent diagnostic of what has actually been stored versus what merely looks familiar.
Mixing subjects within a study block feels harder than finishing one subject at a time, and produces better discrimination between similar rules. The bar tests subjects shuffled together, so at least some of your practice, especially question sets, should be mixed rather than blocked.
Placing items along a mental walk through a familiar building genuinely works for ordered lists, which is why memory athletes use it. The catch is cost: building and maintaining palaces for several hundred rules is a project in itself. If a particular ordered list refuses to stick, a small palace is a fine special-purpose tool. As your primary system, it is a beautiful way to spend hours you do not have.
Drawing the structure of a subject can genuinely help you see how doctrines relate, and the drawing itself is generative work. Just be honest about what it trains: organization, not recall. A mind map session counts as understanding work. It does not replace a single retrieval rep.
Put together, the evidence-backed bar prep stack looks like this: understand the material once through your main course, convert it into element-level rule statements, drill them with retrieval on a spaced schedule, attach a mnemonic to the few genuinely list-shaped rules, blurt each subject weekly as a diagnostic, and interleave your practice questions. Nothing exotic, nothing expensive, and every hour lands on a technique with evidence behind it.
Because re-reading builds recognition, not production, and it decays in days. Switch the time into retrieval practice on a spaced schedule and the forgetting problem shrinks fast.
Retrieval practice, spaced over time. It is the most supported finding in the research and it matches exactly what the essays demand.
For ordered lists, yes, but they are expensive to build at bar scale. Most takers do better spending those hours on retrieval and spacing.
The retention stack, built in: rule statements drilled with active recall and spacing, from $69 →