Most current guides on contraception cite this Guttmacher study for its results on the pull-out method, also called coitus interruptus. That study was performed back in 2009, and found that with perfect use, the pull-out method had a failure rate of about four times out of 100. Not bad, really! But, of course, the whole problem with the pull-out method is that it's sort of hard to perform it perfectly every single time. By averaging that failure rate with the rate of pregnancy when it was used incorrectly all the time, the researchers landed on a failure rate of about 18 percent as a best guess for average use of the method. That's only slightly worse than the average failure rate of condoms (17 percent), though it's pretty easy to argue that condoms are easier to use than the pull-out method.