It starts with an age-old question: If a man pulls out before ejaculating, can a woman still get pregnant?

In bedrooms, basements and the backs of cars worldwide, millions of sexually active humans make choices (or regret them) based on what should be foundational fertility knowledge. Most trusted sources say the answer is yes—it is unlikely but possible that pregnancy will occur, so don’t risk it.

Dig deeper, though, and it quickly becomes unclear exactly where the risk is coming from. Instead of evidence-based education, you’ll encounter some of the most durable misconceptions in sexual and reproductive health. When researchers analyzed a year’s worth of questions that were submitted to an emergency contraception Web site, they found that almost half of the questions that involved sexual acts “express fear about the pregnancy risk posed by pre-ejaculatory fluid.”

Preejaculate—which pretty much everyone calls precum—is the lubricative secretion that is emitted, involuntarily, from the Cowper’s gland in the penis during sexual arousal. Its job is to create a hospitable ride for sperm that ultimately pass through the urethra during ejaculation. But whether you query the Internet or an andrology expert about the fertilizing power of that egg-white goo, you’re likely to get an answer to a different question—that is, a declaration that pulling out is a terrible form of birth control.

“When we’re talking about what’s in preejaculate, that’s not really the point,” said Michael Eisenberg, director of male reproductive medicine and surgery at Stanford University School of Medicine, after I’d asked him the fertilizing-power question in various ways. “We know that pulling out is not effective at preventing pregnancy.”

The pullout method—alternatively known as “withdrawing” or “pull and pray” and formally christened in Latin as “coitus interruptus”—is an ancient form of contraception. The Talmud refers to it as “threshing inside and winnowing outside.” Globally, it is still one of the most commonly used forms of birth control, particularly in regions without access to modern methods. When performed perfectly every time, it actually has a failure rate that isn’t much higher than that of condoms: 4 percent versus 2 percent, respectively. That means about four out of 100 women who rely on the pullout method exclusively will become pregnant during one year of use.

But real life is rarely perfect. Some males cannot reliably perceive the imminence of ejaculation and withdraw too late. Others might emit semen intermittently or over a long period of time instead of as a single event, according to a 1970 family-planning manual. A lot of men don’t realize that the highest concentration of sperm occurs in the first spurt of semen—which can be especially problematic if getting drunk slows down their reaction time. Still others don’t pull out in time because their pleasure takes precedence over a woman’s health and well-being. For reasons such as these, the “typical use” failure rate of coitus interruptus jumps to between 20 and 30 percent.

People in the reproductive-health field largely dismiss the pullout method because they don’t believe men have the ability and willpower to withdraw at the correct time, every time. Meanwhile there is a shocking lack of research on whether or not viable sperm are actually present in preejaculate.

The best way to synthesize the answers I collected from physicians, peer-reviewed journals and educational institutions is this: Preejaculate itself does not contain sperm—or maybe it does occasionally, but perhaps it gets contaminated with sperm that has “leaked” from elsewhere. Plus, there’s leftover sperm from previous ejaculation. And anyway, Eisenberg says, we should assume that preejaculate “usually has some sperm, which can lead to [contraception] failure.”

It is obvious to blame inadequate sex education for our collective confusion. But ironically, write the authors of a 2009 Contraception paper, “the notion that pre-ejaculatory fluid can cause pregnancy ... seems to have been introduced by the medical profession itself.”

Dispelling a Myth?

Where did the fertile prowess of preejaculate originate? Perhaps it was in 1931, when Abraham Stone—a physician and colleague of Planned Parenthood founder Margaret Sanger—wondered how it was even possible for the withdrawal method to fail: Sperm are made in the testicles and don’t route through the Cowper’s gland on their way out. Stone asked some buddies with microscopes to examine their preejaculate for sperm. Among the 24 samples from 18 men, only four contained many or a few sperm. In a 1938 book, Practical Birth-Control Methods, Stone wrote that these figures were insignificant. Regardless, a “myth” that a handful of sperm in preejaculate makes coitus interruptus unreliable took off, and it was “copied uncritically from one textbook another,” according to the 1994 edition of the book Fertility Control.

This myth was popularized by the classic 1966 textbook Human Sexual Response, by William H. Masters and Virginia E. Johnson, according to the Contraception paper. These pioneering sex researchers “warned of the possibility of pregnancy from withdrawal due to the presence of sperm in secretions of the Cowper’s gland”—a statement that “was apparently not evidence-based but subsequently repeated,” the authors write.

The Contraception paper’s authors also speculate on why sperm seem to have “extraordinary potency” in the eyes of the public. In textbooks and the media, sperm are “often anthropomorphized as masculine, forceful, competitive, and single-mindedly determined to fertilize the egg against all obstacles,” they write. Indeed, the memorable 1989 educational film The Making of Me features cartoon sperm “men” in a literal race for a sexualized egg “woman,” set to a soundtrack that includes Richard Wagner’s “Ride of the Valkyries.” Additionally, girls often learn to be terrified of sperm yet aren’t taught how their own body works: A recent survey of 1,000 American women of reproductive age found that 80 percent of them were not able to correctly answer how many days of each cycle they are fertile.

Since Stone’s experiment, there has been little incentive to research coitus interruptus at all, partly because unlike condoms or intrauterine devices (IUDs), there’s no contraceptive product to sell. While the pregnancy risk of preejaculate has only been investigated a handful of times, the results challenge popular assumptions and raise new questions.

Here’s what the literature tells us: In the early 1990s, a study examined the preejaculate of HIV-positive men to determine if the virus was present. (It was.) An ancillary but “more significant” finding described in Contraceptive Technology Update was that “most pre ejaculate samples did not contain any sperm and those that did had only small clumps of a very small amount of sperm which seemed to be immobile.” If a larger study confirmed the results, the article said, it “may dispel the myth that pre ejaculate fluid contains sperm.”

Only tiny studies have taken place since. In a 2003 experiment with 12 Israeli men who gave at least two samples of preejaculate each, scientists examined the secretions under a microscope and found that none of them contained sperm. Another small study also found no sperm.

Several years ago, researchers in England and the U.S. set out to more rigorously investigate the fertilizing potential of preejaculate, noting that “no study has found motile sperm in the pre-ejaculate.” Their paper, published in Human Fertility in 2011, analyzed 40 samples of preejaculate from 27 volunteers. Ten of the volunteers (37 percent) produced samples that included “a reasonable proportion” of motile sperm.

Because some of the men gave samples on multiple separate occasions, an intriguing pattern emerged: sperm was present in either all of an individual’s samples or in none of them. “It would appear from our study,” the authors wrote, “that some men repeatedly leak sperm in their pre-ejaculatory fluid while others do not.”

They therefore concluded, “it is tempting to speculate that the use of withdrawal as a means of contraception might be more successful in some men because they are less likely to release sperm with their pre-ejaculate.”

Then, in 2016, a larger study of 42 healthy Thai men reported that “actively mobile sperm” were found in only 16.7 percent of the subjects. Unfortunately, the researchers did not collect preejaculate samples on multiple occasions.

To make sense of these conflicting data, I called John Amory, a physician and professor at the University of Washington, who is known for his research on male infertility and novel forms of contraception. I asked him about the plausibility of this “two groups” concept: the idea that men might either always have sperm in their preejaculate or never have it.  

Amory responded with surprise. “See, I didn’t even know that,” he said about the studies. “We were taught [in medical training] that sperm were left over from the last ejaculate.” This is a popular theory. Planned Parenthood similarly says that preejaculate “may pick up sperm from a previous ejaculation as it passes through a man’s urethra.” Wikipedia promotes a familiar fix: just urinate before intercourse, the logic goes, and you’ll flush out lingering sperm.

Though the acidity of urine does harm sperm, I could not find any evidence to prove that this strategy is solid. In fact, researchers in the 2011 Human Fertility paper wrote that the volunteers giving samples had, of course, gone to the bathroom several times since their last ejaculation. Therefore, every time the authors observed sperm in preejaculate, the contamination “must have taken place immediately prior to ejaculation.” Clearly, there are consequences to misunderstanding this facet of male fertility.

“Fertility Is a Team Sport”

Because we know so little about sperm in preejaculate, the failure rate of pulling out is really more of an “educated guess” and a topic of controversy among experts in the field. The reality is that lots of people in the U.S. use this method to avoid pregnancy. So, do males approach withdrawal as a serious form of contraception and take responsibility for learning how to maximize its efficacy? While plenty of men feel confident discussing the minutia of abortion and female reproductive parts they tend to be quite ignorant of their own fertility.

Greg Sommer discovered just how few males understand their contribution to pregnancy when he launched an at-home sperm-testing kit called Trak. In 2017, he brought his product to the Consumer Electronics Show in Las Vegas. “We had a demo kit at our booth, and I can’t tell you how many guys came up and said, ‘So, what, I pee in the cup?’”Sommer recalls. “And we had to tell them, ‘No, there’s no sperm in your urine.’”

Sperm awareness got a boost in 2017, when a meta-analysis showed that sperm counts of men from the U.S., Europe, Australia and New Zealand had dropped by more than 50 percent in less than 40 years. “Men are responsible for nearly half of infertility cases but take way too long to get a semen analysis when they are not conceiving naturally,” Sommer says. The study was widely framed as a potential crisis in male fertility, sparking some men to consider their sperm functionality more deeply—or just consider it at all.

Whereas women have long shouldered the burden of both preventing pregnancy (with drugs) and causing pregnancy (with assisted-reproduction technologies such as egg freezing), “there is a growing understanding that fertility is a team sport,” Eisenberg says. “We need to understand more about the male side.”

Recent population surveys have shown that many men do want more birth-control options. Without contraception methods beyond condoms, vasectomy and withdrawal, some guys are already doing “all sorts of crazy and potentially dangerous things to make themselves less fertile to avoid pregnancy,” Sommer says.

In discussion forums on Trak’s infertility education Web site at Sommer found that some men “are biohacking themselves” by using prescription steroid creams to intentionally squash sperm count. Others sit in a hot tub every day. One guy wrote about his “hacked-up underwear heater-type device with a little battery pack,” Sommer says. “Don’t underestimate men’s drive and creativity when it comes to having a better sex life”—meaning men will indeed make efforts and take risks to have sex without condoms.

The Center for Male Contraceptive Research & Development even exploits this incentive to solicit volunteers for clinical drug trials. One image on the center’s Instagram account features a boxer with a punching bag. “Done with condoms? Join the fight for male birth control,” it reads, followed by the hashtag #LoveWithoutTheGlove. It seems to be working: A major clinical trial for a hormonal gel began late last year.

It sounds woefully apropos that scientists and entrepreneurs are convincing guys to learn about reproductive responsibility by appealing to their sexual pleasure—particularly at a time when some U.S. lawmakers want to investigate the “criminality” of miscarriages and classify treatment for ectopic pregnancy as an “abortion.”

Yet more options and knowledge for preventing pregnancy are good things for everyone. After all, nearly half of all pregnancies in the U.S. are unintended, and the lack of access to birth control and health care providers is not the only problem. Nearly 40 percent of women are not satisfied with the birth-control method they are currently using, according to the Guttmacher Institute. When people dislike their contraception for whatever reason—including health side effects from the pill or the tactile compromises of condoms—they are less likely to use it correctly and consistently.

One day, if the pharmaceutical industry decides to reverse course and fund the development of innovative birth control, we could get genetic tests and other technologies to help people of both sexes figure out what kind of contraception might work best for our individual physiologies and ways of life. In addition to hormones and IUDs, researchers could investigate “proteins, enzymes and genes involved in the reproductive process that could be targeted to prevent pregnancy in both women and men—and potentially do so in more precise ways,” wrote journalist Maya Dusenbery in the May issue of Scientific American.

With a personalized-medicine approach, imagine if birth control could be catered to the specific needs and priorities of an individual. In some cases, the task of preventing pregnancy could be truly shared between a couple. “What if the male partner is willing to take on some of the risks and side effects to lower the risks and side effects of his female partner?” Amory says. “No one has really talked about the idea of reframing risk paradigms.”

Until this equitable future arrives, understanding the fertilizing potential of an individual’s preejaculate could give some men another way to participate in the responsibility of contraception. Let’s say that males do fall into two groups, as the Human Fertility study speculates. What if a man—my boyfriend, for instance—could undergo a preejaculate sperm evaluation?

If so, my boyfriend and I might scientifically resolve the final variable in our birth-control efficacy. We use coitus interruptus1 during my fertile window, the weeklong span during which his sperm can potentially fertilize my egg. (An egg is only viable for fertilization for up to 24 hours per menstrual cycle, and sperm can survive in the female body for up to five days.) I determine this window using a technique called the symptothermal method, a means of avoiding pregnancy that involves meticulously tracking changes in cervical fluid and basal-body temperature in order to predict, and then confirm, when ovulation occurs.2

We devised this contraception strategy based on our personal risk-benefit analysis and combined physiologies—and it has worked for us so far. But I’d prefer to empirically validate the absence (or problematic presence) of sperm in my boyfriend’s preejaculate. Frustrated by the paltry research, I decided to conduct an experiment myself.

For Science!

The Trak test, while approved by the Food and Drug Administration, is not designed for testing preejaculate. Nor is it intended to be used as form of pregnancy prevention. But according to Sommer, it is sensitive enough to pick up on sperm concentration as low as one million per milliliter (M/mL). While that sounds like a lot, “the chance of pregnancy is extremely low,” Amory says. “In fertility settings, we take care of a lot of men with those counts who never conceive spontaneously.” The World Health Organization has determined that suppressing sperm counts to this threshold appears to decrease the chances of conception to less than 1 percent per year.

I ordered a Trak fertility kit and recruited one study participant: after assuring my boyfriend that his genetic material wouldn’t be sent off to a lab and end up in a database (Trak isn’t connected to the Internet), he gave me his informed consent.

First, we did a control test to get a sense of his sperm baseline. After 48 hours of abstinence (the minimum length of time for proper semen analysis, according to the WHO), he proffered a five-milliliter ejaculation sample. Per the instructions, we let it sit for 30 minutes to liquefy, gave it a good swirl, then deposited a pipette’s worth of fluid into a test prop. That went into the Trak “engine,” an adorably sized, battery-powered centrifuge.

My boyfriend stared down the engine until it beeped to signal its finish, recalling the way women glare at pregnancy tests while awaiting the results. A white column in the prop reached above the 55 M/mL mark, signaling that his sperm concentration made it into the “optimal” range for conception. After another 48 hours of abstaining from ejaculation (“for consistent science,” I insisted), it was time to test his preejaculate.

“I think accurately testing just precum might be a challenge,” Sommer wrote when I informed him of my plans to use his test for off-label endeavors. “Collecting a sample via masturbation might have different discharge dynamics than during intercourse.”

The hallowed pages of Scientific American are not the place to describe how we collected a full milliliter of unadulterated preejaculate. I will say that our methodology was informed by the science of arousal, a commitment to rigorous research standards and an abundance of humor.

Per the discussions of methodology in the academic studies, we knew it was critical to collect only preejaculate. The authors of the Thai paper wrote that study volunteers might have smeared semen on the collection slides instead of preejaculate, which could mean the number of preejaculate samples that were found to contain sperm was artificially high. In other words, the subjects might have been sloppy, leading to false positives.

(Anecdotally, appealing to male pride created a strong motivation for my volunteer to endure the 30-ish minutes it took to retrieve enough volume of pure preejaculate to run the Trak test. “Wow, look at how much you’re producing,” I cheered about halfway through. By comparison, the academic study subjects were likely masturbating, presumably alone, in a lab, and I humbly hypothesize that they may have gotten bored. The authors of the 2011 Human Fertility study even suggested that subjects might have knowingly handed over samples of ejaculate fluid because they were embarrassed they couldn’t produce sufficient preejaculate.)  

We ran the preejaculate test just as with my boyfriend’s ejaculate: a full pipette of well-mixed fluid went into the prop, followed by a six-minute spin in the centrifuge. Then we peered into the measuring strip under bright light and couldn’t find even a speck of white. If there was sperm present, the concentration was likely below one million per milliliter, which means my boyfriend’s preejaculate sample could be considered infertile by WHO standards.

Though promising, one at-home test doesn’t confirm anything. We would need to replicate this experiment several more times. Sperm count in semen changes over time and is affected by health factors, so perhaps the same is true for preejaculate. Because Trak is not intended for such diagnostics, it would be best to compare the results of our experiments with lab tests at a fertility clinic (if they’d even indulge such a request).

Larger questions abound: Even if there are sperm in preejaculate, can they swim? Are all of their parts intact? And if the sperm present in preejaculate aren’t simply “left over” from the last ejaculation, then from where might they be “leaking,” as the literature suggests?

Filling these knowledge gaps has the potential to fine-tune the math of pregnancy risk. Imagine if males were able to better gauge whether the pullout method is a useful tool in their contraception arsenal or, more critically, whether it is too risky even when the act itself is performed correctly every time.

After all, contraceptive use in the real world is more varied and circumstantial than the behavioral patterns that determine “failure rates.” Few people use only one method in the same exact way every time they have sex. Recent surveys suggest that coitus interruptus is actually employed more frequently than previous research suggests and often in conjunction with other methods. If some men do consistently have viable sperm in their preejaculate, it might help explain the 4 percent failure rate of the withdrawal method despite “perfect” use. It would not be the first time the medical field was wrong to blame contraceptive failure on user error instead of physiological variation.

At the least, researching the mechanisms of preejaculate and pregnancy risk could add evidence-based nuance to sex education. As Amory told me after reviewing the studies on preejaculate, “I think this is an example of when you drill down on a ‘truth,’ one finds it’s not based on much.”

1. We could use condoms during my “fertile window,” but their failure rate over time is not significantly lower than coitus interruptus. Given the best available science and our personal considerations, we chose to be in control over preventing user error rather than risk the uncertainty of product failure.

2. The symptothermal method should not be confused with the rhythm method or similar counting techniques. With perfect use, it can be just as effective as the pill at preventing pregnancy. While I chart my data in a cycle-tracking app, I do not consult predictive algorithms to determine when I am fertile. Like all contraceptive methods, the symptothermal method is certainly not right for everyone. It can, however, be used as an excellent educational tool for learning about fertility and reproductive health