A doorway can end one event and begin another
The familiar failure is small but oddly complete. You leave the sofa to fetch scissors, reach the kitchen, and find that the goal has vanished. The room is visible. The intention is not.
Calling this a memory wipe makes the episode sound more dramatic and less accurate than it is. A doorway does not delete the thought. Instead, crossing into a different place can change the mental context used to retrieve it. The intention remains in memory, but it is no longer attached to the situation receiving most of your attention.
Psychologists describe part of this process using event models. At any moment, the brain maintains a working representation of what is happening: where you are, who is present, which objects matter, what you are trying to do, and what should happen next. The model is selective because working memory cannot keep every recent detail equally available.
In the living room, the unopened package and the need for scissors belong to the same event. Walking into the kitchen introduces a refrigerator, a sink, different objects, and different possible actions. If the brain treats that transition as a boundary, it updates the model. The scissors intention loses the support it received from the previous scene.
The first experiments compared distance with boundaries
The doorway effect became prominent through experiments by Gabriel Radvansky and colleagues. Participants moved through virtual rooms, picking up objects and placing them elsewhere. The researchers could compare travel within one room with a similar journey that crossed into another room.
After crossing a doorway, participants were sometimes slower or more likely to make errors when asked about a recently carried object. Further work reproduced a location-updating cost with smaller screens and in real rooms. It also tested a tempting explanation: perhaps memory suffered only because the room at recall differed from the room at encoding.
That account did not fully work. When participants crossed a boundary and then returned to the original room, restoring the location did not simply restore performance. The transition itself appeared to matter. Once the mind had divided the experience, going back through the door did not stitch the laboratory episode together again.
This result does not mean retracing your steps is useless in ordinary life. A real room contains much richer cues than an experimental space. Returning to the sofa may reveal the package, the movement you were making, or the conversation that produced the intention. Those details can reconstruct the missing chain even if the doorway effect is not reversed in a narrow object-recognition task.
The effect is real enough to study, not universal enough to become a rule
Popular explanations often stop at “doorways reset short-term memory.” The research is less tidy.
A 2021 study tested the effect across immersive virtual reality, video, imagined movement, and walking through real rooms. Most of its experiments did not find the expected simple memory impairment. Under heavier working-memory load, a doorway-related difference appeared mainly as more false alarms rather than the clean deletion suggested by the popular version.
Other methods still find boundary effects. A 2026 virtual-building study tested how people remembered the order of objects encountered across five rooms. Order memory was better for objects from the same room than for objects separated by a doorway. That fits the broader idea that an event model binds information within an episode more tightly than information spread across episodes.
The safest conclusion is conditional. Doorways can influence memory, but the size and form of the effect depend on the task, memory load, similarity of the rooms, and how strongly the transition is experienced as a new event. Everyday forgetting also adds several uncontrolled variables: standing up, changing direction, losing sight of the cue, noticing a phone, or encountering a fresh task in the destination room.
Event boundaries solve a larger memory problem
If boundaries sometimes make information harder to retrieve, why use them at all? Because an undivided record of experience would be difficult to understand.
The brain does not store a day as one continuous file. It recognizes episodes such as making breakfast, commuting, holding a meeting, or searching for scissors. Changes in location, goal, time, topic, or activity help mark where one episode ends and another begins. Information within an event becomes linked; different events remain distinct.
That organization carries a trade-off. A detail from the previous episode may lose immediate priority, but structured episodes are more useful than an indiscriminate stream. Research on event cognition finds that boundaries guide attention and influence how long-term memories are organized. They help separate actions that belong together from actions that happened nearby but served a different goal.
The doorway effect is therefore not evidence that the memory system is broken. It is an awkward consequence of a system sorting continuous life into usable parts.
Make the intention travel with you
The most reliable countermeasure is to give the goal a cue that survives the transition. “Get the scissors from the kitchen” is more portable than “do the thing.” Repeating the specific action keeps it active while the surroundings change.
A physical cue is stronger still. Carrying an empty mug to the kitchen makes the next action visible. A short note works for the same reason: paper does not need working memory. You can also pause before moving and picture the action in the destination room, linking the new setting to the goal before you arrive.
If the intention disappears, avoid filling the gap with a second task. Mentally replay what happened just before you stood up, or return to the object that prompted the trip. The useful cue may be the room, but it may instead be the package on the sofa or the last sentence of a conversation.
Doorways are only one version of the problem. Opening a browser tab, switching apps, ending a call, or changing topics can all create event boundaries without a physical door. The stage changes, a new goal takes control, and the previous intention waits offstage for the right cue.
The brain has not reset. It has organized the scene slightly faster than the unfinished task could follow.