I have used this method very nearly exclusively with all my long-term partners over the last 15 years; the only two times I have gotten pregnant (on the first cycle, no less) were planned non-withdrawals.
I think this method gets a bad rap, in general. It is reported to be 96% effective when used perfectly, according to
this table from Contraceptive Technology. It may be worth noting that while most perfect-use rates have been clinically evaluated, the one for withdrawal -- like the one for some forms of FAM -- is an estimate based on
clinical expertise. Contraceptive Technology also states withdrawal's "typical use" effectiveness rate as 73%. You can read more about how this typical use rate is determined from the
Guttmacher Institute.
There tends to be a lot of debate about whether or not pre-ejaculate contains sperm, but pre-ejaculate is a fluid that does not originate from the testicles and that is used primarily for lubrication of the male's urethra to prepare for the release of semen. Semen and sperm are produced elsewhere in the testicles.
While scientific studies discussed
here and
here show that
on its own, pre-ejaculate has not been shown to viable sperm, the possibility remains that, in real life encounters, small amounts of sperm may be present in a male's pre-ejaculatory fluid. In most cases this would be sperm that was left in the urethra from a previous ejaculation. While it is not often the case, it is still a possibility that should not be taken lightly if one wants to avoid pregnancy. It's also worth noting that the sample sizes in these studies were small, which may limit the conclusions we're able to draw about them. As one of the abstracts explains, "[a] larger study is needed to verify these results."
The important thing is to have a partner who is capable of withdrawing before he has reached ejaculatory inevitability, and for the male partner to urinate between ejaculations in order to flush the urethra of lingering sperm. You can read more
here.