You may not think twice—or feel the least bit guilty—when you grab a little 100-calorie pack of Goldfish crackers. After all, it’s only 100 calories, right?

Not so, according to Richard Wrangham, Ph.D., a human evolutionary biologist at Harvard University. As he told “Discover” last month, “the calorie counts we’ve grown so used to consulting are routinely wrong.

QUIZ: Measure Your Diet

To understand why they are wrong, we have to travel back to the turn of the 20th century, when Wilbur Atwater, Ph.D., and his colleagues at Wesleyan University developed a system for calculating the energy that people gain from eating different foods. In a nutshell, the Atwater method takes the total energy of a food and subtracts from it the energy of the part we don’t digest, thus calculating the total amount of energy we obtain from a food, or how much of that food’s energy is bioavailable.

From these calculations, we get the commonly accepted calorie values for different types of food molecules: nine calories per gram for fats and four calories per gram for proteins and carbohydrates. These standard values are used to calculate the calorie counts on nutrition labels.

However, researchers believe the Atwater system could be vastly improved. “The problem with the in-minus-out approach is that it ignores two important costs of the digestive process,” explains Ph.D. candidate Rachel Carmody, one of Dr. Wrangham’s graduate students, who studies the effects of cooking on food metabolism. “First, the human gut contains a large population of bacteria, and those bacteria metabolize some of our food for their own benefit. The current measurement system does not discriminate between food that is digested by the human versus the bacteria. Second, it doesn’t account for the energy spent digesting food, which can be substantial.”

When we calculate calories, we assume that nearly all fats, carbohydrates and proteins are somewhat equal, although the system does allow for different values for things like animal-derived versus plant-derived protein. But the nine calories per gram for fats, for example, was calculated using olive oil, which has a very different composition of fat molecules than, say, butter. What’s more, various types of proteins, fats and carbohydrates are digested differently by our bodies (which is why we assign 0 calories to Splenda, even though it is technically a carbohydrate).

MORE: Calories On Restaurant Menus Understated

The standard calorie counts also don’t take into consideration how the food was prepared. This oversight is huge: “Food processing externalizes part of the digestive process, making it easier for the body to extract a larger proportion of calories from the same ingredients,” explains Carmody.

The notion that cooking affects how much of a food we digest might seem strange, but it comes as no surprise to food chemists. They have been showing that cooking affects the bioavailability of starch-rich foods such as potatoes for a while now, although how it affects meat has yet to be determined.

When food chemists looked at the total starch content of raw and cooked Matooke flour, for example, they found that the mice fed a cooked flour diet gained two grams over a 21-day period, while their counterparts who were fed the raw flour diet lost a gram. Carmody and her colleagues wanted to examine this phenomenon in depth and find out whether it was the cooking itself that led to differences in absorption or the physical processing involved in cooking like pounding and grinding.

So they fed mice diets of meat or sweet potato that differed in their level of processing: raw and whole, raw and pounded, cooked and whole or cooked and pounded. Carmody and her colleagues found that while pounding the foods did increase the amount of energy the mice obtained from their diet, it was negligible compared to the affect of cooking. Mice fed cooked diets were heavier than those on raw ones—period.

So why does cooking matter? Well, first off, applying heat to a substance alters it physically. We bake potatoes, for example, because heat changes the hard, inedible root into a softer form that is much more palatable—a process known as gelatinization. But gelatinization doesn’t just change how soft and tasty the starch is. It also breaks apart the tightly packed carbohydrate molecules, making them more accessible to our body’s digestive enzymes.

Cooking also involves chopping, grinding and other physical changes to food that increase the surface area, which allows enzymes greater access. “Greater exposure to acids and enzymes results in more food being digested by the end of the small intestine, before it comes into contact with bacteria,” explains Carmody. “Due to its ability to induce both physical and chemical changes in food, the effects of cooking exceed those of non-thermal processing methods like pounding or grinding… In other words, processed diets allow us to absorb more calories while expending less.”

Dr. Wrangham and Carmody hope that their research will prompt the food industry to reconsider how they label caloric content on foods. But they both note that further research is necessary to determine exactly how much different cooking methods change calorie counts.

When it comes to our diet, the potential for weight loss by eating just raw food might seem tempting. After all, as Carmody points out,  “all things equal, incorporating more raw items into the diet would likely promote weight loss.” But she warns, “there are risks in relying too heavily on raw foods.”

Studies of raw foodists have found them to experience higher than expected rates of chronic energy deficiency, so severe as to interfere with ovarian function. “From an evolutionary biologist’s perspective, such data suggest that the caloric gains conferred by cooking may be necessary for normal biological function,” says Carmody.

Until our nutrition labels better align with the science of digestion, the best thing you can do is eat a diet of both raw and cooked foods. If you’re going to count calories, make sure you take into account how much that food has been processed. Not all calories are created equal.

QUIZ: What Kind of Eater Are You?