Our Methodology

Background

There is no current government agency tasked with making a public list of Concentrated Animal Feeding Operations (CAFOs), and as a result there does not currently exist a complete accounting of the number of animal agriculture facilities. This leaves an informational gap for regulators to set policy, legislators to set law, and individuals to make personal decisions.

Factory Farm Watch exists to make the most complete list of animal facilities possible, primarily through combining disparate sources and carefully validating information to make the most complete and up-to-date database of animal agricultural facilities.

The information presented on FactoryFarmWatch.org is compiled from publicly available records, third-party sources, and independent research. While we strive for accuracy, we cannot guarantee that all data is complete, current, or error-free

Methodology

There exist two primary public datasets managed by government agencies which have the most facilities, as well as animal count data for the years released. The California Integrated Water Quality System Project (CIWQS) maintains a database of facilities in order to monitor their potential impacts on water quality; and the California Air Resources Board maintains the California Dairy Database (CADD). The animal count numbers in these databases make up the majority of Factory Farm Watch's listed animal counts.

California Integrated Water Quality System Project (CIWQS)

The CIWQS dataset tracks numbers of animals and water quality violations. It's released every year, and the dataset for 2025 was used. The dataset contains a list of 2,094 facilities, with 134 of those being listed as zero animals, and 204 determined to be duplicate facilities.

California Dairy Database (CADD)

The CADD dataset was created to track cattle populations and digester adoption through 2022. It does not have guaranteed future releases but has released an updated version which goes up to 2023, which is not yet integrated. Because cattle are the primary source of methane from the animal agriculture sector, the dataset tracks exclusively beef and dairy operations, and additionally breaks down animal count by age of the animal. It contains a list of 2096 total facilities, 1,876 of which are dairies and 220 of which are beef. The average breakdown of ages is shown below

Agreement

These two largest datasets of facilities in California had a high level of agreement about animal counts across facilities present in both listings, which gives good confidence that their numbers are comparable and consistent in type. Not all facilities were listed by both sources - CADD had more dairies and cattle facilities than CIWQS, but does not track any other types of facilities unlike CIWQS. The chart shows the breakdown of overlap and non-overlap between the facilities; as well as manually found facilities that were not listed in either source.

Although there was generally good agreement on the numbers of animals at each facility, there were some disagreements. These disagreements are attributable primarily to differing reporting levels between CIWQS and CADD; with CIWQS publicly reporting the number of only one classification of cattle present, as well as being across different years. 209 facilities had an exact agreement on animal count; Figure D below shows a graph of the correlation between the two datasets. Because the CADD dataset reports the full number of animals present at the facility, Factory Farm Watch opts to use the CADD number when there is a disagreement. Factory Farm Watch is working to get access to the full numbers, as this likely undercounts the number of cattle from more recent years and where no CADD data is available.

The names of facilities were often not identical between facilities. In these cases Factory Farm Watch attempted to preserve the most recent name as best as could be determined, though many farms have different names across different databases and their business labels on maps.

Manual Research

Many facilities are not tracked by any government agency. Factory Farm Watch used a manual review to add facilities based on satellite data in order to improve completeness. Additionally, 91 facilities were listed without animal counts in the CIWQS or CADD datasets. This nearly always represents a closed facility, and Factory Farm Watch was able to confirm 78 of these facilities as closed and find numbers for one still open facility. From a combination of satellite, in-person, and online research, we were also able to add 145 facilities which were fully absent from both datasets. 27 of these facilities have had animal counts sourced via online listings or news reports within the past 5 years.

Brand and Processor Relationship Methodology

Identifying the brands and processors connected to individual factory farms was the largest and most labor-intensive manual research effort in the project. There is no comprehensive public database that links animal agriculture facilities to the consumer brands they supply, nor do most brands publicly disclose the specific farms they source from. As a result, supply-chain relationships are often opaque and highly complex: a single facility may sell animals or products under dozens of different brands, while a single brand may source from hundreds of farms across the state or beyond.

To address this gap, Factory Farm Watch relied on a multi-method manual research process, cross-checking independent sources and only recording relationships when there was clear evidence of a connection. As with any manual research, it is impossible to guarantee a connection or that supply chains have not changed since information was gathered.

Avian Flu

Avian flu information was gathered from a number of sources, and carefully matched to facilities in the dataset. Disbursement data for avian outbreaks is tracked and made available at govspending.org, and Farm Forward provided cattle outbreak disbursement information from documents received from Freedom of Information Act Requests (FOIA). The Animal and Plant Health Inspection Service (APHIS) also lists confirmed cases.

Emissions

Emissions calculations were done using the International Panel on Climate Change (IPCC)'s Chapter 10: Emissions from Livestock and Manure Management (IPCC 2006a). The Emissions Factors (emissions per animal per year, hereafter EF) were calculated using the IPCC Tier 1 approach, and used to translate animal populations into methane emissions estimates per facility. This was then multiplied by the IPCC Sixth Assessment Report's (AR6) equivalent global warming potential factor for methane over a 20-year time horizon, 79.7, to reach its listed CO2 equivalent emissions.

The IPCC distinguishes enteric (digestive) and manure emissions. Enteric emissions are only applicable to cattle, and their emissions factors are found in IPCC Table 10.11; facilities classified as having dairy cows used the 'Dairy' EF and remaining facilities used the 'Other Cattle' EF.

Manure emissions for most animals are available in IPCC tables 10.14 and 10.15; and are average temperature based. The annual temperature was based on the ERA5 European Centre for Medium-Range Weather Forecasts (ECMWF), which was accessed September 2025 and contains data until 2023. The 2023 average for the temperature 2m band, which contains temperatures two meters above the ground, was used for the facility's temperature value. The manure emissions factor was then found using this temperature number.

You can download the data here.