Willpower, like a bicep, can only exert itself so long before it gives out; it's an extremely limited mental resource.

Given its limitations, New Year's resolutions are exactly the wrong way to change our behavior. It makes no sense to try to quit smoking and lose weight at the same time, or to clean the apartment and give up wine in the same month. Instead, we should respect the feebleness of self-control, and spread our resolutions out over the entire year. Human routines are stubborn things, which helps explain why 88% of all resolutions end in failure, according to a 2007 survey of over 3,000 people conducted by the British psychologist Richard Wiseman. Bad habits are hard to break—and they're impossible to break if we try to break them all at once.

Some simple tricks can help. The first step is self-awareness: The only way to fix willpower flaws is to know about them. Only then can the right mental muscles get strengthened, making it easier to succeed at our annual ritual of self-improvement.

The brain area largely responsible for willpower, the prefrontal cortex, is located just behind the forehead. While this bit of tissue has greatly expanded during human evolution, it probably hasn't expanded enough. That's because the prefrontal cortex has many other things to worry about besides New Year's resolutions. For instance, scientists have discovered that this chunk of cortex is also in charge of keeping us focused, handling short-term memory and solving abstract problems. Asking it to lose weight is often asking it to do one thing too many.

In one experiment, led by Baba Shiv at Stanford University, several dozen undergraduates were divided into two groups. One group was given a two-digit number to remember, while the second group was given a seven-digit number. Then they were told to walk down the hall, where they were presented with two different snack options: a slice of chocolate cake or a bowl of fruit salad.

Here's where the results get weird. The students with seven digits to remember were nearly twice as likely to choose the cake as students given two digits. The reason, according to Prof. Shiv, is that those extra numbers took up valuable space in the brain—they were a "cognitive load"—making it that much harder to resist a decadent dessert. In other words, willpower is so weak, and the prefrontal cortex is so overtaxed, that all it takes is five extra bits of information before the brain starts to give in to temptation.

This helps explain why, after a long day at the office, we're more likely to indulge in a pint of ice cream, or eat one too many slices of leftover pizza. (In fact, one study by researchers at the University of Michigan found that just walking down a crowded city street was enough to reduce measures of self-control, as all the stimuli stressed out the cortex.) A tired brain, preoccupied with its problems, is going to struggle to resist what it wants, even when what it wants isn't what we need.

There's something unsettling about this scientific model of willpower. Most of us assume that self-control is largely a character issue, and that we would follow through on our New Year's resolutions if only we had a bit more discipline. But this research suggests that willpower itself is inherently limited, and that our January promises fail in large part because the brain wasn't built for success.

Everybody knows that the bicep has practical limitations: If we ask the muscle to hold too much, it will give out and drop everything on the floor. And just as our muscles get tired after a tough workout, and require a rest to recuperate, so does the poor prefrontal cortex need some time off.

In a 2002 experiment, led by Mark Muraven at the University at Albany, a group of male subjects was asked to not think about a white elephant for five minutes while writing down their thoughts. That turns out to be a rather difficult mental challenge, akin to staying focused on a tedious project at work. (A control group was given a few simple arithmetic problems to solve.) Then, Mr. Muraven had the subjects take a beer taste test, although he warned them that their next task involved driving a car. Sure enough, people in the white elephant group drank significantly more beer than people in the control group, which suggests that they had a harder time not indulging in alcohol.

The implications of this muscle metaphor are vast. For one thing, it suggests that making lots of New Year's resolutions is the wrong way to go about changing our habits. When we ask the brain to suddenly stop eating its favorite foods and focus more at work and pay off the Visa…we're probably asking for too much.

The willpower-as-muscle metaphor should also change the way we think about dieting. Roy Baumeister, a psychologist at Florida State University who has pioneered the muscle metaphor, has demonstrated in several clever studies that the ability to do the right thing requires a well-fed prefrontal cortex.

In a 2007 experiment, Prof. Baumeister and his colleagues found that students who fasted for three hours and then had to perform a variety of self-control tasks, such as focusing on a boring video or suppressing negative stereotypes, had significantly lower glucose levels than students who didn't have to exert self-control. Willpower, in other words, requires real energy.

In another experiment, Mr. Baumeister and his colleagues gave students an arduous attention task—they had to watch a boring video while ignoring words at the bottom of the screen—before asking them to drink a glass of lemonade. Half of the students got lemonade with real sugar, while the other half got a drink with Splenda. On a series of subsequent tests of self-control, the group given fake sugar performed consistently worse. The scientists argue that their lack of discipline was caused by a lack of energy, which hampered the performance of the prefrontal cortex.

Since the most popular New Year's resolution is weight loss, it's important to be aware that starving the brain of calories—even for just a few hours—can impact behavior. Skipping meals makes it significantly harder to summon up the strength to, say, quit cigarettes. Even moderation must be done in moderation.

The final piece of the willpower puzzle is distraction. Research by Walter Mischel at Columbia University and others has demonstrated that people who are better at delaying gratification don't necessarily have more restraint. Instead, they seem to be better at finding ways to get tempting thoughts out of their minds.

For instance, Prof. Mischel has found that four-year-old children who are better at resisting the allure of eating a marshmal low—they get a second marshmallow if they can wait for 20 minutes—are the ones who sing songs, play with their shoelaces or pretend the marshmallow is a cloud. In other words, they're able to temporarily clear the temptation out of consciousness. (Prof. Mischel has also shown that these "high delayers" go on to get higher SAT scores and have lower body-mass indexes as adults.) Because they know that willpower is weak, they excel at controlling the spotlight of attention: When faced with candy, they stare at the carrots.

While this willpower research can get dispiriting—the mind is a bounded machine, defined by its frailties—it also illustrates some potential remedies. Prof. Baumeister figured that it might be possible to strengthen willpower by exercising it, and in 1999, he asked a group of students to improve their posture for two weeks. Interestingly, these students showed a marked improvement on subsequent measures of self-control, at least when compared to a group that didn't work on sitting up and standing straight.

The lesson is that the prefrontal cortex can be bulked up, and that practicing mental discipline in one area, such as posture, can also make it easier to resist Christmas cookies. And when a dangerous desire starts coming on, just remember: Gritting your teeth isn't the best approach, as even the strongest mental muscles quickly get tired. Instead, find a way to look at something else.

—Jonah Lehrer is the author of "How We Decide" and "Proust Was a Neuroscientist."