Estimating liberally to include not only "legitimate" rapes but also unsuccessful perceived attempts compiled by advocacy groups with the stated purpose of stamping out sexual assault, 1 in 6 women will be victimized at some point during their lives. Not all of these will occur during the period of a woman's life when she's fertile, but with the aim of overshooting if anything, for the sake of argument assume they do.
There are a few days every four weeks when a healthy woman aged ~14-45 is capable of becoming impregnated. About 75% of these fertilizations something will go awry--sperm will lose its way, the egg will shut sperm out, or the body will reject the fertilized egg before or after implantation, etc. One-quarter of the time, a baby will be on the way.
From the CDC, about 4 in 5 sexually active women (according to the GSS, about 1 in 10 women in this age range are not sexually active at any time during the year) of reproductive age use pill-based contraception. Again for sake of argument, disregard other less widely used methods like depo-provera and the today sponge.
So, starting from the top: 1/6 = .1667, 3/28 = .1071. Multiply these to get .0179, and then multiply this by .25 (the successful one-fourth of fertilizations). We get .0045. Finally, multiply this by .28 (the one-fifth of women not using the pill from the 90% of at least marginally sexually active women plus the one-tenth who, not being sexually active, are presumably not on birth control, ignoring that it has other non-procreative uses) to get .0013, or 0.13% of women--1 in 750, with liberal estimation parameters--for whom the question of "legitimate" rape will be a pregnancy-inducing concern.
Getting that out there is worth forfeiting the opportunity to retake the Senate, isn't it?
GSS variables used: AGE(18-45), YEAR(2000-2010), SEX(2), SEXFREQ