Raw Milk Illness Rate Per Consumer: Updated Data from 2022
A per-consumer illness rate needs two numbers: how many people get sick, and how many people drink raw milk. The illness estimate most often used comes from Costard et al. 2017. The consumer estimate behind it comes from a survey conducted in 2006 and 2007. A 2022 study by FDA researchers provides a newer consumer count, and sales data since then suggest the consumer base has grown further.
The older consumer estimate
Costard et al. 2017, published in Emerging Infectious Diseases, estimated that outbreaks linked to unpasteurized milk and cheese cause about 730 illnesses a year in the United States (96% of an estimated 761 dairy outbreak illnesses). For consumption, it used the FoodNet Population Survey from 2006 to 2007, which found that 3.2% of the population drank raw milk. Applied to the US population of about 300 million at the time, that is roughly 9.6 million people, and a per-consumer illness rate of about 1 in 13,000 per year.
Earlier versions of this calculation circulated as “1 in 14,250,” based on a rounded 10 million consumers and the 761 figure for all dairy. The more precise inputs give about 1 in 13,000.
The 2022 FDA estimate
Lando et al. 2022, published in the Journal of Food Protection by researchers at FDA’s Center for Food Safety and Applied Nutrition, drew on two nationally representative FDA surveys: the 2016 Food Safety Survey and the 2019 Food Safety and Nutrition Survey. It found that:
- 4.4% of US adults drank raw milk at least once in the past year, about 11 million people
- 1.6% drank it once a month or more, about 4 million
- 1.0% drank it once a week or more, about 2.5 million
Consumers were more likely to be younger, live in rural areas and live in states where retail sale of raw milk is legal.
Dividing Costard’s 730 annual illnesses by 11 million consumers gives about 1 in 15,000 per year. This combines figures from different studies, time periods and survey methods, so it is an approximation rather than a finding of either paper. The reasonable range across the two consumer estimates is about 1 in 13,000 to 1 in 15,000.
The per-consumer rate and the consumer counts behind it are summarized in raw milk statistics.
Growth since 2019
A 2025 survey by the Annenberg Public Policy Center found that 4% of US adults drank raw milk in the past year, in line with FDA’s 4.4%. It did not ask how often. Sales data point upward: retail raw milk sales rose from $12 million in 2021 to $19.4 million in 2022, according to Raw Milk Institute figures reported by Ambrook, and weekly raw milk sales in 2024 ran about 21% above 2023, as reported by the University of Delaware. Several states have also expanded legal access since the 2019 surveys.
Past-year surveys have held near 4%, so any growth is more likely among regular drinkers, which a past-year question does not capture. For every national estimate side by side, see How Many People Drink Raw Milk in the US?
Has illness grown with consumption?
If illness rose in step with sales and legal access, reported outbreak illness would be climbing. Through the most recent peer-reviewed data, it has not:
| Period | Confirmed raw milk outbreak illnesses | Average per year | Source |
|---|---|---|---|
| 2005 to 2020 | 1,696 | about 106 | Stephenson et al. 2024 |
| 2009 to 2014 (milk and cheese) | 648 | about 108 | Costard et al. 2017 |
| 2013 to 2018 | 675 | about 112 | Koski et al. 2022 |
These are confirmed outbreak cases, before any adjustment for underreporting, so they are lower than Costard’s modeled 730. They are useful for comparing one period with another because they are counted the same way each time. Single large outbreaks can still move a year’s total: a 2023 to 2024 Salmonella outbreak traced to one California producer sickened 171 people on its own.
What this calculation does not capture
- Sporadic illness. Costard’s estimate counts outbreak-related illness only. Illness outside recognized outbreaks is excluded, as it is for most outbreak-based estimates.
- Mismatched time periods. The illness model uses 2009 to 2014 outbreak data; the newer consumer counts are from 2016 and 2019.
- Differences between farms. The rate is an average across every kind of source, from licensed and routinely tested dairies to unlicensed ones.
For the per-serving measure and the frequency breakdown, see Raw Milk Illness Rate Is Low Per Serving Consumed and Per Consumer. For comparisons with other foods, see Comparing Raw Milk Foodborne Illness Data with Other Foods.
