By Jean Hofve, DVM

A note from Dr. Jean: I wrote this for The Whole Dog Journal around 2001. I put it online after the 2007 pet food recall scandal, so you can understand how such a thing could happen, and who’s in charge of fixing it. It’s updated as of 2012.


While most pet guardians are certain that “someone” is in charge of regulating the manufacture of commercial dog food in this country, very few people know who that mysterious official or agency might be. But somebody’s gotta be making sure that pet food doesn’t contain any harmful ingredients and does contain what animals need to survive, right? The FDA? USDA? Someone?

Unfortunately, the answer isn’t as clear-cut as, “Yes, it’s all taken care of.” There are numerous government and industry agencies that oversee various aspects of pet food production, but there really is no single office that provides seamless overall supervision of the industry. So is there anyone making sure that a “duck and potato” food really contains ducks and potatoes? Or testing the food to see whether it really contains a minimum of the 20 percent protein it claims in its “Guaranteed Analysis”? Maybe, depending on where you live. There are many opportunities for pet foods to fall between the cracks of testing and enforcement. A walk through the many halls of pet food regulation reveals why a reliance on some branch of the government to ensure a food is “nutritionally complete and balanced” is pure folly.


You might rightly assume that the Federal government has some sort of control over the production of food, even pet food. The Food and Drug Administration is charged with the enforcement of the Federal Food, Drug, and Cosmetic Act. A division of the FDA called the Center for Veterinary Medicine (CVM) is responsible for regulating animal drugs, medicated feeds, food additives and ingredients, and pet food, making sure that they conform with the Food, Drug, and Cosmetic Act. This Act requires that pet foods contain no harmful substances, and be truthfully labeled. However, only in extreme cases does the FDA or the CVM get involved in an investigation of a food maker, and generally, only as a last line of enforcement. The most meaningful regulation of pet foods occurs at the State level.

Each individual state has its own regulations and its own Department of Agriculture (or similar department, in a few states), which oversees the sale of pet food within its borders. Before a new brand or a new type of dog food can be sold in a given state, the maker is required by law to register the new food in each state in which it will be sold. The state’s feed control officials (FCOs) are responsible for examining the food’s label claims and the food itself. Some states have very proactive FCOs, who aggressively examine and test new foods being sold or made within their states’ borders. Kentucky, Texas, and Minnesota, for instance, have a reputation for thoroughness when it comes to testing pet foods.

Most states at least have a program to look at pet food labels. Yes, the size, shape, and color of the fonts and panels, as well as what they say, are regulated. Pet food companies can’t make unsubstantiated claims, nor can they make health or treatment claims without going through several hoops, including the FDA.

What do the states test for? The main area of focus is the Guaranteed Analysis (GA), which the states require to be printed on every container of pet food. The states can (and most do) test for everything that is included in the GA. The only things that are required to be in the GA include the minimum percentages of crude protein and crude fat, the maximum percentage of crude fiber, and the total percentage of moisture in the food; that’s all. Some pet food companies include more information in their GA, adding minimum levels of certain vitamins, fatty acids, or other nutrients they believe the consumer will appreciate. This is going out on a limb for the manufacturers, because it just about guarantees that at least some states will test for these items as well. Pet food companies that ship nationwide tend to stay within these guidelines.

Before we discuss other tests or standards a pet food might be held to, we have to introduce another organization, one that influences the states’ policies on pet food.


Many people have heard of the Association of American Feed Control Officials (AAFCO), and assume that this is the agency that polices the pet food industry. However, AAFCO has no regulatory power whatsoever; it can neither approve nor ban foods. Rather, AAFCO is a non-governmental, non-regulatory, voluntary organization of feed control officials (FCOs) from each state (as well as several other countries). Its role is advisory. AAFCO exists to provide a forum for discussion by all interested parties, address issues of quality and standardization for animal feed and pet food, provide nutritional standards for pet foods and guidelines for feed and pet food manufacturing and labeling, and outline enforcement actions for the real regulators at the state level, the FCOs.

AAFCO influences the production of pet foods only as far as those states that have adopted its “model” regulations in whole or in part. (Some states have adopted their own standards, which may or may not be similar to AAFCOs.) AAFCO is the place where the state feed control officials can go to discuss issues of feed safety, animal health, and interstate commerce with the all the other people who know the most about these issues. Then they go home and set and/or enforce policy for their states.

In order to obtain the information about every imaginable aspect of pet food, AAFCO allows experts from many different fields to join the conversation, as it were, to educate the FCOs in the finer details of associated specialties. Formerly called “liaisons,” these advisors do come from the pet food industry, including the Washington DC lobbying group Pet Food Institute, as well as the grain and feed industries, the rendering industry, laboratories, farm co-ops, and other groups with an interest in AAFCO’s decisions. The Animal Protection Institute (now BornFree USA), for instance, had an advisory position on the Pet Food Committee and Ingredients Definitions Committee for several years participate in AAFCO subcommittees as advisors; there are committees on feed safety, ingredient definitions, state/industry regulations, and many more. Currently, there are two advisors serving on behalf of pets and consumers–myself (representing the Pet Welfare Alliance), and Susan Thixton from Committees and working groups may be formed to deal with a specific issue, or dissolved when their work is done or the issue is no longer relevant. (Click here for more on AAFCO from the industry’s perspective.)

The presence of so many vested parties, all of whom would like to influence the feed control officials to benefit their own aspect of the industry, worries many animal welfare activists, and some even regard AAFCO as a sort of pawn of industry that does not have our animals’ health at heart. However, only the state feed control officials and the FDA and USDA representatives, are voting members of AAFCO. At AAFCO meetings, which are held twice a year, advisors often speak on issues where they have an interest or stake in the outcome. Advisors’ comments are considers by the FCOs and then the issue is voted on by the FCOs. In my experience, the FCOs, as a group, are definitely not pro-industry; they take their role as watchdog very seriously, and it is not a case of “the fox guarding the henhouse” as some have claimed.

For example, around 1999 or 2000, as more and more consumers were getting on the internet and learning about commercial pet food, the rendering industry pushed to have the official feed term “by-products” re-named “animal proteins.” This was debated in the Ingredient Definitions Committee (IDC). The proposal was turned down because the IDC felt it was anti-consumer, and  the FCOs believed that the new term was being requested, not because of a change in the ingredient itself, but merely to obscure and confuse the issue for consumers.

In spite of that defeat, the renderers approached the IDC not long afterward with another request, this time to change the name “poultry by-products” to “poultry and bone meal.” As an advisor to the IDC from the Animal Protection Institute, I argued strongly against this change, as did representatives from two major pet food companies and others. (FYI, they were Nutro and Purina, in case you want to give them credit.) The IDC voted unanimously against the change.

Most people do not know that AAFCO allows anyone to attend its meetings and to speak or ask questions from the audience. There is a significant registration fee (more than $400 for the January 2013 meeting), as well as travel costs, but attendees are not screened or restricted.


Perhaps AAFCO’s biggest legacy to the state FCOs as well as consumers has been the development of two tools for the standardization of pet food formulation. Most, but not all, states have adopted these development tools, and require foods sold in those states to adhere to one of the two AAFCO standards. But, remember, it is the state FCOs who check to see whether the food makers are compliant with the standards.

1. Nutrient Profiles. The first standard is the AAFCO Nutrient Profiles, an effort to identify the minimum (and a few maximum) levels of “macronutrients” (protein, fat, and fiber) and the “micronutrients” (vitamins, minerals, essential amino acids) that research has shown to be necessary for dogs and cats. Years ago, pet food makers manufactured foods to the nutritional standards set by the National Research Council, there were numerous faults with those standards. AAFCO convened panels of experts in canine and feline nutrition to develop new, better standards. These were adopted by AAFCO around 1990 (a little earlier for dogs, a little later for cats).

Although the Nutrient Profile system has done a lot to standardize the business of pet food production, it’s not without its critics. There are studies that suggest some nutrient levels may be too high, and others too low. The Nutrient Profile system of formulation does not address the issue of ingredient quality whatsoever. One critic of this method (Dr. Mark Morris, who also founded Hill’s Pet of feed formulation designed a “food” that met all the AAFCO nutrient profile requirements – even though the food was primarily formulated from old shoe leather, sawdust and motor oil with a multi-vitamin-mineral supplement. Obviously, there would be no guarantee that any animal would eat such a food, or could digest it, even though it contained all the vitamins, minerals, protein, fat, etc. that the nutrient profiles required.

The Nutrient Profiles were based on the recommendations of the National Research Council (NRC) from the 1980s. The NRC published new recommendations in 2006; AAFCO has adopted new standards that will be published in 2014.

2. Feeding Tests.
The second method for pet food formulation addresses those concerns – but contains some loopholes, as well. AAFCO developed a protocol for a six-month feeding trial that can be used as a tool to determine whether a food can sustain life in a target test population (dogs or cats in all life stages, or specific stages of growth of maintenance). (The growth/lactation protocol is only 10 weeks, but they have to run more extensive blood tests to analyze it.) The test population is fed nothing but the food in question for six months, and if the subjects test normal (on weight and a few blood tests), the food passes. This method at least would help a maker demonstrate that the food is palatable and digestible enough to maintain life in the test population – something the nutrient profile system doesn’t do. And this method is good if a feed maker has some brilliant research that indicates the levels of certain nutrients in the AAFCO nutrient profiles are inadequate for promoting maximum health, and they can formulate a food that they think is better; and they can conduct feeding trials to prove their food works. At present, some large breed puppy foods are being feed tested because the companies believe that the Nutrient Profiles don’t address their special requirements.

However, the feeding test requires only eight test subjects, and only six must finish the trial. Many nutritional deficiencies or overdoses would not appear in this short period; the feed’s fitness for maintaining longevity, reproductive, or multi-generational health would not be demonstrated. (In reality, most pet food companies using feeding tests use many more animals than the 8 required, and the major manufacturers do conduct generational studies.)

These two systems necessarily miss a lot of potential problems. A food meeting the Nutrient Profile may or may not pass a feeding trial; and foods that have passed a feeding trial don’t have to meet all specifications of the Nutrient Profiles. Clearly, it would be possible for a marginal food to pass these tests, yet fail to provide adequate nutrition in the long run, and in fact such problems are well documented. In generational studies, where animals were kept on the same food for three to five generations, researchers at the University of California at Davis found that some foods that pass feeding trials still won’t support animals over the long term. They estimated that, of 100 foods that pass AAFCO analysis criteria, 10 to 20 would not pass the feeding trials, and of those, 10 percent would not be adequate for long-term feeding. A former FDA nutritionist emphasizes, “The formulation method does not account for palatability or availability of nutrients. Yet a feeding trial can miss some chronic deficiencies or excesses.”

In the case of minimum requirements without a corresponding maximum, some foods contain significant nutrient excesses that may actually be dangerous in the long run. The Kentucky feed control officials analyzed test data from all pet foods tested during 1994 and 1995, and found that certain nutrients, such as magnesium, iron, and manganese, were present in most dry dog and puppy foods at 200-400 percent or more of their AAFCO Nutrient Profile values. Their conclusion: the AAFCO profile for certain nutrients is not a reasonable indicator of the actual level present in many products. An excess of many minerals, including copper, magnesium, iodine, and iron, may produce signs of toxicity over time. Excess iodine, for instance, is thought to be a possible factor contributing to the epidemic of hyperthyroidism among older cats. As it turns out, manufacturers deliberately and routinely add 300-400 percent of many vitamins and minerals to ensure that by the end of the “best by” period (12-18 months in many cases), the food still meets minimum requirements.

And there is one more big wrench in the works: manufacturers are allowed to test one food of a similar “family” of foods, and apply that certification to all foods in that family. There is no way for the consumer to know from the label which foods were actually tested by feeding trials certification, making that “gold standard” more like fool’s gold than the real thing.

If a food has met either AAFCO requirement, it may state on the label that the food is “complete and balanced.” These label statements are why many people are under the mistaken impression that AAFCO actually regulates the food industry, or is a governmental agency. Neither is true.

Remember, it’s the states that are in control – but they control only the pet food manufacturers who want to sell food within their borders. Only the state’sFCOs have the ability to approve or deny the right of a manufacturer to sell a particular food in their state, or to punish manufacturers for labeling infractions. And the only way they can make these decisions is to test the various foods that the makers register for sale there.


As mentioned above, some states test only  the Guaranteed Analysis information (crude protein, crude fat, fiber, moisture). Others test individual nutrients (amino acids, vitamins, minerals) as well. Nearly every manufacturer has had one or more foods fail one or more tests at one time or another. Many foods fall short on the stated protein levels. Some are too high in ash or fiber. Even more ominous is the failure of tests for major minerals such as phosphorus or calcium. The manufacturers assert that tests on any particular batch or lot of food may not be representative of all their foods, but because such failures are so widespread, from the cheapest generic to the most specialized and expensive foods, it is a very disturbing trend. The fault often lies with the “premix” that supplies vitamins and minerals to the food. A horrifying example of this occurred in 2011, when many cats eating Wellness canned foods became ill or died from thiamine deficiency; and the same problem caused another widespread recall, this time from Diamond Pet foods, in March 2013.

State feed control officials can and do enforce violations of their states’ regulations, but this process is not sweeping and not always swift. Depending on the nature of a problem they discover with a food, there are numerous levels of notification and correction; in the mean time, tons of non-compliant food can be sold and consumed by our pets. Each state compiles an annual report which lists the violations; these documents are public record. Many states publish this data; a few post it on the Internet.

Regulation and enforcement of the pet food industry varies widely from state to state. Some states have adopted very tough legislation, and others have minimal pet food laws. Some states scrutinize foods carefully, and others hardly at all. And you can’t assume any coordination between the state’s regulatory aims and its follow-through on enforcement. California, for example, has one of the nation’s most restrictive pet food production Acts in the country, the “Pure Pet Food Act of 1969.” It prohibits 4D meat and other bad stuff in pet food. However, the Act isn’t enforced as intended; both FDA and AAFCO have taken the position that these products are fine as long as they stay out of the human food chain.

In reviewing the states’ reports, it’s obvious that every food fails something, somewhere, some time. But the most striking trend is that the foods with the most problems tend to be locally produced and regionally marketed; there are numerous small pet food companies that make foods that are sold in one state only or across one state border only. The national manufacturers stick closer to the rules; if they ship nationally, they pretty much have to make their products to whichever state standards are the strictest.


If you’re like me, the more you learn about the pet food industry, the more you worry about your pets’ food! The kind of regulation and oversight that many of us assume is present over the industry as a whole really doesn’t exist. Instead, existent regulation and the vicissitudes of the market itself tends to promote the better products, and weed out the “bad actors” over time. It really is amazing that the industry is as “clean” as it is – but this isn’t, perhaps, saying much. In an ideal world, every food in the country would have to pass feeding trials and lab tests that prove sufficient (and not harmful) nutrient levels on an ongoing basis. But in this world, our pets represent the “test animals,” and we are providing the feeding trials.