In 2016 a Rutgers University laboratory dropped 2,560 food samples on the floor. The headline result is short: some bacteria move onto the food instantly, in under a second. The five-second rule does not hold.
The interesting part comes after that verdict. In the same study, contact time mattered less than what you dropped and what you dropped it on. A slice of watermelon and a gummy sweet on the same tile pick up wildly different numbers, and that gap is wider than the gap between one second and five minutes.
What was actually measured
The design matters, because it tells you how to read the numbers. Four surfaces: stainless steel, ceramic tile, wood and carpet. Four foods: watermelon, bread, bread with butter and gummy candies. Four contact times: under a second, five seconds, thirty and three hundred.
A 25 cm² patch of each surface was inoculated with 1 ml of Enterobacter aerogenes and dried for five hours, leaving roughly 10 million cells on the surface. Food with a 16 cm² contact area was dropped from a height of 12.5 cm.
Contact time, food type and surface type all affected transfer, all three at high statistical significance. The authors’ own conclusion is blunt: longer contact does give more transfer, but the nature of the food and the surface matter as much or more.
Why nobody tested this until recently
Before 2016 the rule had almost no scientific backing. The Rutgers authors list what existed at the time: one published study, at least two television programs, and results from two university laboratories that circulated as press releases and were never printed.
That single published study was Clemson University’s 2007 paper, whose title contains the phrase “testing the five-second rule”. Everything before it rested on common sense and TV. That is the usual life of a kitchen rule: it spreads faster than anyone can test it, and it survives for decades after the test is done.
Moisture beats seconds
Watermelon collected the most, anywhere from 0.2% to 97% of everything on the surface depending on conditions. Gummy candies collected the least, 0.1% to 62%. Bread ran from 0.02% to 94%, bread with butter from 0.02% to 82%.
The spread within each food is enormous, which honestly reflects reality: one dropped slice proves nothing. The ranking, though, is stable. A wet cut surface works like blotting paper, a dry sweet picks up almost nothing.
Viruses behave the same way. A Dutch group at Wageningen measured norovirus transfer and got 7% of infectivity onto cucumber slices against 0.3% onto whole tomatoes, and gave the same explanation: the cucumber slice carries more moisture at the contact surface.
For viruses the argument “they need time” collapses entirely. A virus on the floor does not multiply. It lies there and waits. What counts is contact, not duration.
Pressure and real contact area
Those design details are doing real work. Rutgers dropped food from 12.5 cm onto a 16 cm² contact area; the Dutch team pressed a finger pad onto surfaces at 0.8-1.9 kg/cm² for about two seconds. Both were measuring the same underlying thing: how tightly the food met the surface.
That makes the watermelon-versus-sweet gap easier to understand. A soft wet slice molds itself to the microrelief of the tile and touches almost all of it. A hard sweet rests on a few points, with air in the gaps. Moisture works together with the area of genuine contact rather than on its own.
The practical version: a slice that lands cut-side down and the same slice landing skin-side down are two different events, even though the stopwatch reads the same.
Carpet gives up less than tile
This is the most counterintuitive result in the field. The Clemson group compared Salmonella transfer onto bologna from tile, wood and carpet back in 2007. Carpet gave up under 0.5%. Wood and tile gave up between 5% and 68%.
The reason is the geometry of contact. Pile holds cells down among the fibers and touches food only at the tips. Smooth tile hands over nearly everything it holds: in the same experiment, over 99% of cells moved from tile to bologna in five seconds of contact.
None of which licenses eating off the carpet. Carpet is dirtier than tile precisely because it accumulates and holds. As an illustration of the main point, though, it is perfect: the surface decides more than the stopwatch.
How long bacteria last on a dry floor
The same Clemson work tested survival. After 28 days, tile still carried 1.5 to 2.5 log CFU per square centimeter, which is tens to hundreds of cells. Salmonella comfortably survives four weeks on a dry surface in numbers sufficient for transfer.
The Dutch work shows the same effect over shorter intervals. Norovirus transfer from finger to stainless steel was 13% immediately after contamination and fell to 0.1% after ten minutes of drying. A fresh puddle on the floor is more dangerous than an old mark in the same spot, and that difference is measured in minutes and hours.
Time on the surface does work in your favor, just slowly. With bacteria sitting on the surface for 2, 4, 8 and 24 hours, transfer fell from 6.5 to 4.8, 4.6 and 3.9 log CFU. A day of drying cuts transfer roughly four hundredfold. Five seconds change nothing.
Why the percentages should not be taken literally
Another paper from the same Rutgers laboratory, 2003, 352 observations. It found that the transfer percentage depends on how many bacteria sat on the source: where there were more, the percentage was lower; where there were fewer, it was higher.
That sounds like a statistical footnote and is anything but. You cannot line up “97%” from one study against “5%” from another, because the difference may live in the starting inoculum rather than in the surface. And 97% in a study with ten million cells on a tile says little about what crosses from your floor onto a piece of bread.
So the useful thing to remember here is directions, not figures. Wet picks up more than dry, smooth releases more than fibrous, a fresh spill beats an old one, and seconds barely register.
What this means in a real kitchen
The question is what was on that floor before your food arrived. Floors do not manufacture pathogens. Pathogens get there from raw meat, shoes, the dog’s bowl and the dishcloth.
Hence the working logic.
- Dry food off a dry clean floor is the lowest risk case here. A cookie, a nut, a piece of candy.
- A wet cut surface, watermelon, tomato, cucumber slice, sliced meat, picks up the most. That piece is better off in the trash.
- Location beats duration. The floor by the trash can, by the door, or where you jointed a chicken an hour ago will not become clean in five seconds. How long bacteria from raw chicken survive on kitchen surfaces is covered in should you wash chicken.
- For children, pregnant women, older people and anyone immunocompromised the rule is simple: dropped means binned. In those groups the infectious dose is reached more easily.
- Food heading for a pan or a pot gets a second chance from cooking. Food heading for a salad does not.
Quick summary
- The five-second rule is disproved directly: some bacteria cross in under a second.
- Moisture decides more than time: watermelon took up to 97%, gummy candies from 0.1%.
- Smooth tile hands over more than 99% in five seconds; carpet under 0.5%.
- Salmonella survives on dry tile for four weeks in numbers that still transfer.
- Percentages from different studies are not comparable: they depend on the starting bacterial load.
- What decides is the food, the surface, and what was on that floor beforehand.
How pathogens reach floors and countertops at home is covered in our piece on kitchen cross-contamination. How few cells it takes to make you ill is shown through infectious dose. Why attached bacteria resist rinsing is covered in our piece on biofilms. What to do with food that thawed sooner than planned sits in our piece on refreezing. Why reheating cannot fix rice that sat warm for too long is explained in our piece on reheating rice. The wider picture is in our food safety hub.
Kitchen beliefs often rest on a sound measurement read as an answer to a different question. The dairy version is covered separately: a market panel against a pathogen test.
Kitchen rules often survive without anyone measuring the size of the effect. The same treatment for the never-refrigerate list: one entry was retired by its own author.
