What are the odds of getting food poisoning from eating food left out of the fridge?
Evidence quality 4.38/5
Eight-dimension review score against the quality rubric . Each dimension scored 1–5.
- D1 Source grounding
- 5/5
- D2 Source authority
- 5/5
- D3 Arithmetic
- 4/5
- D4 Uncertainty
- 4/5
- D5 Scope
- 4/5
- D6 Prose
- 5/5
- D7 Perception honesty
- 3/5
- D8 Caveat completeness
- 5/5
● your factors — click this risk ▾ to reveal
- Your factors
≈ As likely as
Perceived
Most people treat "left out for a few hours" as a dial they can read by smell and appearance. If the chicken looks fine and the rice doesn't taste odd, it gets eaten. The gap between that heuristic and the actual microbiology is large in both directions: the casual per-meal risk is lower than a cautious reader might guess, but the lifetime probability of eventually eating a temperature-abused meal and paying for it is far higher than almost anyone assumes. No cross-national survey isolates this fear cleanly, so the perceived side is editorial intuition rather than polled data.
Rough estimate: Most adults assume the per-event risk is either near-zero ('it's probably fine') or near-certain ('you'll get sick')
Source: editorial intuition, not polled
Actual
~1 in 35 per year (US residents, temperature-abuse-linked foodborne illness)
US residents, all ages, foodborne illness where improper holding temperature was a contributing factor
Show derivation
Starts from CDC's ~48 million domestically acquired foodborne illnesses per year (Scallan et al. 2011, restated on CDC's current food safety page). The CDC's National Outbreak Reporting System analysis for 2014-2022 finds that "allowing foods to remain out of temperature control" contributed to roughly 10-15% of outbreaks during food preparation and another 9-14% during service or display, with "improper cooling" implicated in another 9-17% depending on period. We take ~20% as a central estimate of the share of US foodborne illness where temperature abuse is a contributing factor (the categories overlap, so this is deliberately conservative). 20% × 48 million ≈ 9.6 million cases per year, or about 2.9% of the US population per year — roughly 1 in 35. Compounded over 59 years of adult life: 1 - (1 - 0.029)^59 ≈ 0.82, or about 4 in 5. The uncertainty band runs from 10% contribution (lifetime ≈ 0.58) to 30% (lifetime ≈ 0.93), which spans most defensible readings of the NORS data and the underlying Scallan illness total.
Caveats: The headline "~1 in 35 per year / 4 in 5 lifetime" is an aggregate of very heter…
The headline "~1 in 35 per year / 4 in 5 lifetime" is an aggregate of very heterogeneous scenarios. Most of the per-year incidence is mild gastrointestinal illness that self-resolves in 24-48 hours, not hospitalization or death; the fatal subset is covered in the separate food-poisoning-death entry and runs roughly 1 in 1,860 lifetime. The risk depends overwhelmingly on four variables: food type (high-protein dairy and meat vs dry/acidic), duration out of refrigeration, ambient temperature (the 2-hour rule becomes a 1-hour rule above 90°F), and host susceptibility. Reheating a leftover does not rescue food contaminated with heat-stable toxins from Staphylococcus aureus or Bacillus cereus — those toxins are formed during room-temperature incubation and survive subsequent cooking. The 4-hour breastmilk window is a documented exception driven by human-milk microbiology, not a carve-out the general FDA rule would support. The NORS contributing-factor coding is also imperfect: an outbreak can be coded with multiple factors, and "improper holding temperature" is typically reported alongside cross-contamination or food-worker issues, so the true share attributable solely to temperature abuse is genuinely uncertain within the 10-30% band used in the uncertainty calculation.
How the risk varies
The headline figure averages across very different situations. Here’s how the probability varies by scenario or context:
1 in 1.7 · 60%
The 2012 staph perlo outbreak in CDC MMWR is the closest measured analogue: 62% attack rate among those who ate the dish. Bennett et al. report median attack rates of 75% (B. cereus) and 87% (S. aureus) in outbreak settings. Reheating does not help because the relevant toxins are heat-stable. Cooked rice and pasta are the classic Bacillus cereus ("fried rice syndrome") vehicle — the emetic toxin cereulide forms during room-temperature holding and survives reheating.
1 in 10 · 10%
Listeria and invasive Salmonella proliferate in temperature-abused food and are the main killers in this subgroup. The per-event serious-illness probability rises accordingly, and the fatality conditional on illness is several-fold higher than in healthy adults.
1 in 50 · 2.0%
The typical case — the reference point the rows above and below vary from. A reheated chicken/rice/dairy dish left out half an afternoon at ordinary room temperature, eaten by a healthy adult. Measured per-event rates are not cleanly reported in the literature; synthesized from the share of US outbreaks with temperature abuse as a contributing factor and typical household exposure frequency.
1 in 1,000
CDC and Academy of Breastfeeding Medicine guidance: freshly expressed milk is safe up to 4 hours at room temperature. Human milk's innate immune components (immunoglobulins, lactoferrin, live leukocytes) meaningfully suppress bacterial growth relative to other foods. Serious illness from properly handled milk within the 4-hour window is rare in the reported literature.
1 in 10,000
Staphylococcus aureus, Bacillus cereus, and most foodborne pathogens require water activity above ~0.85 and pH above ~4.6 to grow. Dry goods and acidic foods sit outside both bounds and do not proliferate at room temperature.
Bar length and shade rank these scenarios against each other, not against other risks. The exact odds are shown beside each.
Related risks
Other risks on similar themes — for exploring related fears.
Undercooked food
What are the odds of getting food poisoning from undercooked meat, fish, or eggs?
Raw meat cross-contamination
What are the odds of getting sick from not washing hands or surfaces after handling raw meat or eggs?
Restaurant food poisoning
What are the odds of being hospitalized from food poisoning after eating at a restaurant?
Decisions this risk informs
Choices that turn on this risk — how people weigh the trade-off.
Drink raw milk vs. pasteurized
% who regret this choice
Action dominates
Action regret 5.0× higher
Pick challenger
Same risk, three time-spans. A single mishandled meal runs about 1 in 50 (2%); across a year of meals, about 1 in 35; over an adult lifetime, roughly 1 in 1.2 (82%). The headline figure on this page is the lifetime one — which is why it sits higher than any single-meal scenario in the breakdown above. They are not in conflict; they count the same risk over different spans of time.
The USDA’s operational rule is that food in the “Danger Zone” between 40°F and 140°F should not sit out for more than two hours, or one hour above 90°F, because bacteria double in number every twenty minutes in that range. That rule is about the hazard; the probability that any given room-temperature meal will actually make you sick is a separate question, and the literature gives a less clean answer. Combining CDC’s ~48 million US foodborne illnesses per year with the National Outbreak Reporting System’s finding that roughly a fifth of outbreaks list temperature abuse as a contributing factor yields about 9-10 million temperature-abuse-linked illnesses per year, or roughly 1 in 35 Americans per year. Compounded over a typical adult lifetime, the probability that a US adult will experience at least one illness traceable to mishandled-temperature food comes out close to 4 in 5 — roughly the same order of magnitude as a lifetime cancer diagnosis, with almost all of the count being mild gastroenteritis rather than anything serious.
What makes this fear unusual is the bimodal gap between per-event and lifetime probabilities. Eat a high-protein leftover that sat on the counter for four hours and the per-event risk is somewhere around 1 in 50 for a healthy adult — much lower than the “you’ll definitely get sick” intuition. Eat that same kind of meal a few times a year for forty years and the lifetime odds of at least one bad episode approach certainty, which is why most adults can recall at least one memorable bout of it. The 2012 MMWR report on a military unit lunch party is the clearest measured analogue of the severe-abuse scenario: a chicken-and-rice dish sat in an unheated oven overnight, was reheated the next day, and produced a 62% attack rate among the forty attendees. Reheating did not rescue the food because staphylococcal enterotoxins are heat-stable, formed during the room-temperature incubation and untouched by the second trip through an oven. This is why the “just heat it up again” instinct fails for exactly the pathogen class most associated with leftover abuse.
The clearest pathogen-specific version of this is the one that went viral as “fried rice syndrome.” Bacillus cereus spores survive cooking, and when cooked rice or pasta then sits at room temperature the bacteria multiply and produce cereulide, a heat-stable emetic toxin that — like the staphylococcal toxin above — is not destroyed by reheating. The operative variable is the number of hours the cooked starch spends at room temperature, not whether it is later fried or microwaved, which is why “but I reheated it” is no reassurance. The illness itself is trivial: vomiting that starts within 0.5 to 6 hours and resolves inside a day. What makes the fear vivid is two documented deaths — a 20-year-old man in Brussels who died after eating B. cereus-contaminated spaghetti, with cereulide measured at 14.8 µg/g in the meal, and a child who died of liver failure after a family ate pasta salad. Both are isolated case reports of the rare severe tail; against the roughly 63,400 US B. cereus illnesses the CDC’s Scallan model estimates each year, the modeled annual death count is zero. The reading matches the rest of this entry: the everyday illness is real and fairly common but minor, and the fatal version is real but case-report rare.
The heterogeneity under the headline does most of the work. Dry and acidic foods (bread, jam, pickles, hard cheese, dried pasta) sit outside the water- activity and pH windows that the main pathogens need to grow, so the 2-hour rule functionally does not apply to them. Freshly expressed breastmilk gets a documented 4-hour carve-out from CDC and the Academy of Breastfeeding Medicine, driven by the immunoglobulins, lactoferrin, and live leukocytes that suppress bacterial proliferation in human milk far more effectively than in any other food. High-protein leftovers left overnight are the other tail: measured outbreak attack rates in staphylococcal and Bacillus cereus intoxications run 75-87%, and for the same abuse scenario the risk of serious illness climbs by a factor of three to five for pregnant women, adults over 65, children under five, and immunocompromised patients — groups for whom Listeria, invasive Salmonella, and Toxoplasma in temperature-abused food produce serious disease at doses a healthy adult would clear asymptomatically. The “1 in 35 per year” headline is an average across all of these; the readers whose actual risk is well above it and well below it tend to be the ones for whom the averaged figure is least useful.
Related tidbits
Food left out too long sickens 82% of people over a lifetime. Cross-contamination from raw meat gets 73%. The kitchen counter is statistically more dangerous than the cutting board.
Claim ledger
Every number below is what each source reported, with the verbatim quote we relied on and how we arrived at our figure. Click any link to verify directly.
-
[1] US Department of Agriculture, Food Safety and Inspection Service — "Danger Zone" (40 °F - 140 °F)
"Danger Zone" (40 °F - 140 °F)- Statistic
Bacteria multiply rapidly between 40°F and 140°F, doubling in as little as 20 minutes; food should not be left out more than 2 hours (1 hour above 90°F)- Excerpt
“"Bacteria grow most rapidly in the range of temperatures between 40 °F and 140 °F, doubling in number in as little as 20 minutes. This range of temperatures is often called the 'Danger Zone.' Never leave food out of refrigeration over 2 hours. If the temperature is above 90 °F, food should not be left out more than 1 hour." ”
- Source data from
- 2024-07-12
- Accessed
- 2026-04-16 · archived copy
- Calculation
- USDA FSIS defines the operational rule the rest of the entry is testing. The 2-hour / 1-hour-above-90°F guideline is the regulatory baseline against which "temperature abuse" is measured in CDC outbreak investigations. Used here as the definition of the exposure, not as a probability estimate.
- Independence
- FSIS guidance is the governing US regulatory standard; its framing of the Danger Zone is reused by FDA, CDC, and state health departments. Not independent of those downstream sources — upstream of them.
-
[2] CDC Morbidity and Mortality Weekly Report (MMWR) Surveillance Summaries — Contributing Factors of Foodborne Illness Outbreaks — National Outbreak Reporting System, United States, 2014-2022
Contributing Factors of Foodborne Illness Outbreaks — National Outbreak Reporting System, United States, 2014-2022See all 3 Likelier entries citing this source →
- Statistic
Among 2,677 foodborne outbreaks 2014-2022, 'allowing foods to remain out of temperature control' during preparation contributed to 9.9-15.2% and during service/display to 8.9-13.6%; 'improper cooling' contributed to 8.8-17.3%- Excerpt
“"Allowing foods to remain out of temperature control during preparation contributed to 15.2% of outbreaks during 2014-2016, 12.2% during 2017-2019, and 9.9% during 2020-2022. Allowing foods to remain out of temperature control during food service or display contributed to 13.6%, 10.4%, and 8.9%, respectively, across the three periods. Improper cooling of food contributed to 9.4%, 8.8%, and 10.9%." ”
- Source data from
- 2025-03-13
- Accessed
- 2026-04-16 · archived copy
- Calculation
- CDC NORS is the canonical US surveillance system for outbreak contributing factors. Summing the three temperature-related factors gives an upper envelope of roughly 30-40% of outbreaks, but the categories overlap (an outbreak can be coded with multiple contributing factors), so we deflate to ~20% as the central estimate for the share of US foodborne illness where temperature abuse was meaningfully involved. Applied to Scallan's 48 million illnesses/year: 0.20 × 48e6 ≈ 9.6 million cases/year, or ~1 in 35 of the US population. Over 59 adult-remaining years: 1 - (1 - 0.029)^59 ≈ 0.82.
- Independence
- NORS draws on the same state and local public-health reporting pipeline as CDC FoodNet and the Scallan et al. estimates; treat as a methodological sibling rather than a fully independent data source.
-
[3] US Centers for Disease Control and Prevention — About Food Safety
About Food SafetySee all 4 Likelier entries citing this source →
- Statistic
CDC estimates 48 million illnesses, 128,000 hospitalizations, and 3,000 deaths per year from foodborne illness in the US- Excerpt
“"CDC estimates that each year 48 million people get sick from a foodborne illness, 128,000 are hospitalized, and 3,000 die." ”
- Source data from
- 2024-01-01
- Accessed
- 2026-04-16 · archived copy
- Calculation
- The CDC headline total provides the denominator for the normalized figure. 48 million illnesses / 330 million US residents ≈ 14.5% of Americans per year experience a foodborne illness from any cause. Multiplied by the ~20% share attributable to temperature abuse (NORS) gives the ~2.9% per year figure compounded in the normalization.
- Independence
- Restates Scallan et al. 2011; not independent of the MMWR contributing- factors analysis, which also uses CDC surveillance data.
-
[4] Clinical Infectious Diseases / Bennett et al. (CDC) — Foodborne Disease Outbreaks Caused by Bacillus cereus, Clostridium perfringens, and Staphylococcus aureus — United States, 1998-2008
Foodborne Disease Outbreaks Caused by Bacillus cereus, Clostridium perfringens, and Staphylococcus aureus — United States, 1998-2008- Statistic
1,229 US outbreaks 1998-2008 from three temperature-abuse-linked pathogens; errors in food processing and preparation reported in 93% of outbreaks; median attack rates 75% (B. cereus), 87% (S. aureus)- Excerpt
“"During 1998-2008, 1229 foodborne outbreaks caused by Bacillus cereus, Clostridium perfringens, and Staphylococcus aureus were reported in the United States [...] Errors in food processing and preparation were commonly reported (93%), regardless of etiology. Contamination by food workers was common only among S. aureus outbreaks (55%). Meat or poultry dishes were commonly implicated in C. perfringens (63%) and S. aureus (55%) outbreaks, and rice dishes were commonly implicated in B. cereus outbreaks (50%)." ”
- Source data from
- 2013-04-16
- Accessed
- 2026-04-16 · archived copy
- Calculation
- Establishes the biological mechanism behind the "left out food" scenario: the three pathogens whose disease is overwhelmingly driven by temperature abuse (they produce heat-stable toxins or form spores that survive initial cooking, then proliferate when the cooked food sits at room temperature). The 87% median attack rate for staph outbreaks is the per-exposure figure once a food has been seriously abused — not directly the headline, but the upper anchor for the "long severe abuse" row of the regional breakdown.
- Independence
- Bennett et al. is a peer-reviewed CDC analysis of Foodborne Disease Outbreak Surveillance System data over a distinct time window from the later NORS 2014-2022 summary; treat as a complementary deeper look at the pathogens most associated with temperature abuse.
-
[5] CDC Morbidity and Mortality Weekly Report — Outbreak of Staphylococcal Food Poisoning from a Military Unit Lunch Party — United States, July 2012
Outbreak of Staphylococcal Food Poisoning from a Military Unit Lunch Party — United States, July 2012- Statistic
22 of 35 interviewed attendees (62% attack rate) became ill after eating perlo held in an unheated oven for ~8 hours overnight; reheating failed to destroy heat-stable enterotoxin A- Excerpt
“"Of the 40 attendees, 35 (88%) were interviewed, of whom 22 (62%) met the modified case definition [...] The pot of cooked perlo then was placed in an unheated oven for approximately 8 hours overnight [...] Staphylococcal enterotoxins are resistant to heat treatment; subsequent rewarming of the perlo for approximately 1 hour the following day did not destroy the heat-stable toxin." ”
- Source data from
- 2013-12-20
- Accessed
- 2026-04-16 · archived copy
- Calculation
- Concrete attack rate for the "long severe abuse" subgroup: a high-protein cooked dish left at room temperature overnight, then reheated, produced a 62% illness rate among those who ate it. This is the anchor for the 0.60 probability in the regional_breakdown row for overnight abuse of a high-risk food. Reheating did not rescue the food because the staph enterotoxin is heat-stable — which is why the "just reheat it" intuition is mechanically wrong for this pathogen class.
- Independence
- A specific outbreak investigation, independent of the aggregate NORS and Bennett analyses. Provides the measured per-event attack rate those aggregate sources cannot.
-
[6] US Centers for Disease Control and Prevention — Breast Milk Storage and Preparation
Breast Milk Storage and Preparation- Statistic
CDC guidance: freshly expressed breast milk can be stored at room temperature (77°F or colder) for up to 4 hours, refrigerated for up to 4 days, frozen for 6-12 months- Excerpt
“"Freshly expressed or pumped milk can be stored: At room temperature (77°F or colder) for up to 4 hours. In the refrigerator for up to 4 days. In the freezer for about 6 months is best; up to 12 months is acceptable." ”
- Source data from
- 2024-06-28
- Accessed
- 2026-04-16 · archived copy
- Calculation
- The breastmilk carve-out is qualitatively different from the general 2-hour rule: CDC allows up to 4 hours at room temperature because human milk contains immunoglobulins, lactoferrin, and live leukocytes that suppress bacterial growth substantially better than most cooked foods. Cited here to support the breastmilk-specific row in the regional breakdown rather than the general headline.
- Independence
- CDC breastfeeding guidance is methodologically distinct from the FSIS Danger Zone framework; the 4-hour limit is derived from human-milk microbiology and Academy of Breastfeeding Medicine research rather than from the USDA general-food model.
-
[7] FEMS Microbiology Reviews (Stenfors Arnesen, Fagerlund, Granum) — From soil to gut: Bacillus cereus and its food poisoning toxins
From soil to gut: Bacillus cereus and its food poisoning toxins- Statistic
The emetic toxin cereulide is resistant to acid, proteolysis, and heat and is not destroyed by reheating food held at room temperature after a first heating; emetic illness incubates in 0.5-6 h and lasts 6-24 h- Excerpt
“"Because cereulide is resistant towards acid conditions, proteolysis and heat, it will not be destroyed by gastric acid, the proteolytic enzymes of the intestinal tract or by reheating foods that have been stored at room temperature after a first heating. [...] The duration of the emetic disease is normally 6-24 h." ”
- Source data from
- 2008-07-01
- Accessed
- 2026-06-20 · archived copy
- Calculation
- The mechanism behind "fried rice syndrome." Bacillus cereus spores survive cooking; when cooked rice or pasta then sits at room temperature the bacteria multiply and produce the heat-stable emetic toxin cereulide. Because the toxin is preformed and heat-stable, reheating kills the bacteria but not the toxin — so the operative variable is hours held at room temperature, not whether the food is later reheated. Supports the reheated-rice paragraph; not a probability input.
- Independence
- Peer-reviewed microbiology review; mechanistically distinct from the surveillance and outbreak sources, which count cases rather than explain toxin chemistry.
-
[8] Emerging Infectious Diseases / Scallan, Hoekstra, Angulo, et al. (CDC) — Foodborne Illness Acquired in the United States—Major Pathogens
Foodborne Illness Acquired in the United States—Major Pathogens- Statistic
CDC/Scallan model: Bacillus cereus causes ~63,400 domestically acquired US foodborne illnesses per year (90% CrI 15,719-147,354), ~20 hospitalizations (0-85), and 0 deaths- Excerpt
“"[Tables 2-3, Bacillus cereus, foodborne, estimated mean annual US burden with 90% credible intervals] Estimated illnesses: 63,400 (15,719-147,354). Hospitalizations: 20 (0-85). Deaths: 0." ”
- Source data from
- 2011-01-01
- Accessed
- 2026-06-20 · archived copy
- Calculation
- Anchors the honest scale of the B. cereus / fried-rice risk. The emetic illness is real and not rare — tens of thousands of US cases a year, almost all unreported because the vomiting is short and self-limiting — but the modeled national death count rounds to zero. The 63,400 figure is all B. cereus (emetic plus diarrheal, not decomposable by syndrome); rice- and pasta-borne emetic illness is a subset of it. Used for scale and the death framing, not added to this entry's headline arithmetic.
- Independence
- The Scallan national burden model shares the CDC surveillance lineage of the NORS and FoodNet sources already cited; treated as a methodological sibling and used here only for the pathogen-specific B. cereus line.
-
[9] Journal of Clinical Microbiology / Naranjo, Denayer, Botteldoorn, et al. — Sudden death of a young adult associated with Bacillus cereus food poisoning
Sudden death of a young adult associated with Bacillus cereus food poisoning- Statistic
Case report: a 20-year-old man in Brussels died after eating pasta (spaghetti) contaminated with emetic Bacillus cereus; cereulide measured at 14.8 µg/g in the meal- Excerpt
“"A lethal intoxication case, which occurred in Brussels, Belgium, is described. A 20-year-old man died following the ingestion of pasta contaminated with Bacillus cereus. Emetic strains of B. cereus were isolated, and high levels of cereulide (14.8 μg/g) were found in the spaghetti meal." ”
- Source data from
- 2011-12-01
- Accessed
- 2026-06-20 · archived copy
- Calculation
- One of the two famous fatal case reports behind "fried rice syndrome." It documents that B. cereus emetic intoxication can kill, but it is a single case report — the kind of isolated event that, aggregated nationally, rounds to the 0 deaths/year Scallan models. Cited as the vivid tail, not as a rate. The widely repeated "left out for five days" detail circulates in popular retellings but is not stated in this record and is not asserted here.
- Independence
- An individual forensic case report, independent of the surveillance burden models; supplies the severity ceiling the aggregate estimates cannot.
-
[10] Journal of Clinical Microbiology / Dierick, Van Coillie, Swiecicka, et al. — Fatal family outbreak of Bacillus cereus-associated food poisoning
Fatal family outbreak of Bacillus cereus-associated food poisoning- Statistic
Case report: a fatal case of liver failure after consumption of pasta salad, in a family outbreak of B. cereus emetic food poisoning, demonstrating the possible severity of the emetic syndrome- Excerpt
“"Bacillus cereus is a well-known cause of food-borne illness, but infection with this organism is not commonly reported because of its usually mild symptoms. A fatal case due to liver failure after the consumption of pasta salad is described and demonstrates the possible severity of the emetic syndrome." ”
- Source data from
- 2005-08-01
- Accessed
- 2026-06-20 · archived copy
- Calculation
- The second classic fatal case report. The abstract's own framing — that B. cereus "is not commonly reported because of its usually mild symptoms," yet can in rare cases be fatal — is the perceived-vs-actual thesis in one sentence: an ordinarily trivial illness with a very rare severe tail. Cited for the tail, not a rate.
- Independence
- Independent Belgian case report from a different outbreak and year than Naranjo et al.; together they establish that the fatal tail exists while remaining case-report rare.