The Old Number Was a Guess
For most of the 20th century, textbooks and popular science sources confidently stated that humans could distinguish about 10,000 different smells. This number appeared everywhere, from introductory psychology courses to perfume industry literature. It sounded precise and authoritative.
It was also essentially made up.
The 10,000 figure traces back to a 1927 paper by two chemists, Crocker and Henderson, who proposed a classification system for odors based on four primary smell qualities. They estimated that humans could distinguish about 10,000 combinations within their system. The number was a rough theoretical estimate, not an experimental finding. But it got into textbooks, and once a number is in textbooks, it tends to stay there for a very long time.
Nobody seriously tested the claim until 2014, when Andreas Keller and Leslie Vosshall at Rockefeller University designed an experiment to actually measure human olfactory discrimination.
The Experiment
Keller and Vosshall created odor mixtures from a pool of 128 different odor molecules. They combined these molecules in groups of 10, 20, or 30 to create complex mixtures, then asked participants to distinguish between pairs of mixtures that shared varying percentages of their components.
When two mixtures shared fewer than about 57 percent of their components, participants could reliably tell them apart. When the overlap exceeded about 57 percent, discrimination became unreliable. Using this discrimination threshold and mathematical modeling based on the number of possible combinations from 128 components, the researchers estimated a lower bound of one trillion distinguishable odors.
The researchers were careful to note that one trillion is almost certainly an underestimate. They used only 128 of the many thousands of odor molecules that exist in nature, and they tested only a limited number of mixture combinations. The true number of distinguishable smells could be orders of magnitude higher.
The study was not without controversy. Some researchers argued that the mathematical extrapolation from a limited experimental dataset to a trillion-scale estimate involved assumptions that might inflate the number. A 2015 critique suggested the true number might be much lower, though still far above 10,000. The scientific consensus now sits somewhere between "much more than 10,000" and "possibly a trillion or more," with the exact number still debated.
How Smell Actually Works
To understand how the nose achieves such discrimination, you need to know a bit about the olfactory system's architecture.
High inside each nasal cavity, there is a small patch of tissue called the olfactory epithelium, about the size of a postage stamp. This tissue contains roughly 6 million olfactory receptor neurons (for comparison, a dog has about 300 million, which is one reason dogs have a much more sensitive sense of smell).
Each olfactory neuron expresses one — and only one — type of olfactory receptor on its surface. Humans have about 400 functional olfactory receptor types (out of roughly 800 receptor genes, about half of which are nonfunctional pseudogenes — evolutionary relics that no longer produce working proteins).
When you inhale an odor, the airborne molecules dissolve in the mucus coating the olfactory epithelium and bind to these receptors. Here is where it gets interesting: most odor molecules activate not just one receptor type but several, each to a different degree. And most receptors respond not to one molecule but to many different ones.
This means that each odor produces a unique pattern of activation across the 400 receptor types. Your brain does not identify a smell by which single receptor fires — it identifies it by reading the entire pattern, like a barcode. The combinatorial possibilities of 400 receptors, each capable of being activated at various levels, create an astronomical number of distinguishable patterns. This is the same principle that allows three types of cone cells in your eyes to produce millions of colors — combinatorial coding is enormously powerful.
Smell and Memory
One of the most striking features of the olfactory system is its direct connection to memory and emotion. The olfactory bulb, which processes smell signals, sends projections directly to the amygdala (involved in emotional processing) and the hippocampus (involved in memory formation). Most other senses are routed through the thalamus first, adding an extra processing step.
This direct anatomical connection is thought to be why smells can trigger vivid, emotionally charged memories more powerfully than sights or sounds. The smell of a specific laundry detergent might transport you instantly to your grandmother's house. A whiff of a particular cologne might bring back a relationship that ended years ago. This phenomenon, sometimes called the Proust effect (after the novelist who famously described a flood of memories triggered by the taste and smell of a madeleine cookie), is not just subjective — it has been confirmed in brain imaging studies showing strong hippocampal and amygdalar activation in response to odor-cued memories.
Individual Differences
Not everyone's nose is equally capable. Olfactory ability varies significantly between individuals due to genetics, age, health, and experience.
Genetics play a substantial role. Variations in olfactory receptor genes mean that some people literally have different receptors than others, causing them to perceive certain odors differently or not at all. About 75 percent of people can smell the compound androstenone (found in sweat and pork), while about 25 percent cannot detect it. Among those who can, some find it unpleasant (like stale urine) while others find it pleasant (like vanilla or sandalwood). These differences are genetically determined.
Age reduces olfactory ability. Smell sensitivity typically peaks in early adulthood and begins declining after age 60. Significant olfactory loss (called anosmia when complete, hyposmia when partial) affects about 25 percent of people over 50. Interestingly, loss of smell is now recognized as an early biomarker for neurodegenerative diseases like Parkinson's and Alzheimer's — another example of how sensory changes signal broader health issues.
Training can improve discrimination. Professional perfumers, sommeliers, and food scientists can distinguish odors that untrained noses merge into a single impression. This is not because they have more receptors — it is because their brains have learned to parse the receptor activation patterns more finely through repeated experience.
An Underappreciated Sense
Humans tend to think of themselves as visual creatures with poor senses of smell, especially compared to dogs. While it is true that our olfactory epithelium is smaller and we have fewer receptor neurons than many other mammals, the "humans have a bad sense of smell" narrative is overblown.
A 2017 review by John McGann at Rutgers University traced this myth to a 19th-century claim by neuroanatomist Paul Broca, who argued that the relatively small size of the human olfactory bulb compared to the rest of the brain indicated a diminished sense of smell. This claim was picked up by Sigmund Freud and others and became conventional wisdom.
In reality, the human olfactory bulb is actually large in absolute terms — larger than that of many other mammals — and humans can outperform dogs in detecting certain odors, particularly some synthetic compounds. The one trillion estimate, if accurate, puts human olfactory discrimination on par with or exceeding what we previously attributed only to keen-nosed animals.
Your nose is doing far more work than you give it credit for.
Related: You Replace Your Entire Skeleton Every 10 Years · You Can't Hum While Holding Your Nose · Why Does One Eye Water More Than the Other?
Written by Helen Russo
Helen covers health, wellness, and food topics. She focuses on evidence-based information and practical advice for everyday life.