When searching for a lemon among many fruits, would smelling lemon make it easier to spot? Even without indicating where an object is, scent might provide a cue about what to look for.
In 2025, Serena Castellotti and colleagues investigated whether a scent matching the target changed performance in a task briefly displaying fruit images. The aim was to determine whether task-irrelevant odor information could enter visual search, and whether its influence differed between people.
Searching for briefly glimpsed fruit while smelling a scent
Twenty-two healthy adults participated: 20 women and two men, with a mean age of 28.7. Tubes at the nose delivered apple, lemon, or strawberry scent, or air without added fragrance. Actual fruit was not placed in front of participants; formulated fruit fragrances were delivered by an apparatus.
After five seconds of scent delivery, a photograph of the target fruit appeared in the screen center for just 0.05 seconds. Then one target and seven other fruit images appeared in a circle for 0.1 seconds. After the pictures disappeared, participants clicked the target’s location with a mouse.
Color saturation was halved and presentation was brief because a pilot experiment with more saturated images and longer exposure had produced accuracy of about 90%. The researchers avoided a task easily solved by vision alone, leaving room for scent to help or hinder performance.
Each person completed 96 searches, with odor–target combinations randomized. Participants were not told before a trial which odor would appear or whether it would match the target fruit. Scent did not indicate the target’s location, and the task did not ask participants to use smell.
Comparing matching, mismatching, and absent scents
There were four conditions: air without added scent; a scent matching the target fruit; a scent matching neither the target nor any other fruit on screen; and a scent matching another, nontarget fruit.
For example, when strawberry was the target and lemon scent was delivered, the researchers distinguished between displays without any lemon and displays containing lemon as an incorrect option. If scent drew attention to its matching fruit, performance might be especially impaired in the latter condition.
Accuracy was higher when scent matched the target than in either incongruent condition. Accuracy in the incongruent conditions was also lower than without scent. However, matching scent did not significantly increase accuracy relative to the no-scent condition.
On correct trials, responses were faster with a matching scent than without scent. Neither incongruent condition differed significantly in response time from no scent. Thus, the comparisons showing differences were not the same for accuracy and speed.
The measured time ran from image onset to the mouse click. It included decision-making and hand movement as well as visual detection. The prediction that a scent-matching distractor would attract more incorrect selections or slow responses more than the other incongruent condition was not supported.
Who benefited most from a matching scent?
The authors also focused on differences between participants. Exploratory, post-experiment correlations suggested that people with lower no-scent accuracy showed larger improvements with congruent scent. Conversely, those performing better without scent tended to show larger declines with incongruent scent.
The authors suggest that matching scent may have helped direct attention for people who found the visual information difficult. Those already performing well with vision alone had less room to improve, while an incongruent odor may have interfered as extraneous information. These are interpretations of individual differences.
This was a post hoc analysis comparing baseline performance with change from that same baseline in the same 22 people, so replication in other participants is needed. It does not establish a general everyday category of people who are poor at finding things and therefore benefit from scent.
Being good at identifying odors was not necessarily an advantage
After visual search, participants also smelled fragrances and selected the corresponding fruit from eight images. Mean accuracy was 73%, with substantial differences between participants.
Lower accuracy on this task correlated with greater improvement from congruent scent during visual search. Better odor discrimination did not necessarily mean more benefit from scent. The authors discuss the possibility that detailed odor processing may sometimes compete with the visual task.
Finally, when asked to name the odors they thought had been presented, 14 of the 22 participants correctly named all three. Naming performance did not correlate significantly with visual-search performance, leading the authors to consider influences that do not require explicit naming of the scent. However, awareness was not measured on each trial, so this does not establish an unconscious effect.
Scent may support attention to the object being sought
The odors did not directly identify the target’s location. Nevertheless, performance changed depending on whether scent and target represented the same fruit. The authors emphasize the possibility that semantic congruency about the type of object being sought supports visual attention.
However, selecting briefly presented images cannot fully separate effects on perception from response biases about what to choose when uncertain. Eye-movement measurements could clarify processes during search. The small sample of 22, predominantly women, also limits generalization.
The study did not test situations such as shopping for fruit in a real store, with time to look around, pick things up, and choose. The authors suggest possible relevance to everyday food choices, while noting that confirmation requires more realistic conditions.
The study received support from DAAD, the German Research Foundation, and grants associated with the EU’s Next Generation EU program, among others. The authors declared no competing interests.
Summary
A scent matching the fruit being sought produced higher accuracy than incongruent scents and faster correct responses than no scent. Meanwhile, the extent of scent’s help or interference appeared to vary with participants’ visual-search and odor-identification performance.
The results suggest that visual search is not exclusively visual: a scent referring to the same object may also influence attention. It is not simply a case of adding fragrance to help everyone search better. What is being sought, how difficult it is to find, and how scent is processed may jointly determine how the senses help one another.
References
Castellotti, S., Soldo, M., Plank, T., Del Viva, M. M., & Greenlee, M. W. (2025). Visual search performance depends on the congruency of olfactory sensations. Scientific Reports, 15, 38116. DOI: 10.1038/s41598-025-25995-1. Full text. The original article is available under CC BY 4.0.
