Abortion can cause infection, hemorrhage, bowel injuries, clotting disorders, and other potentially fatal complications. Also, amniotic fluid, fetal tissue, or air can get into the woman's bloodstream and cause an embolism, which can be fatal.
The degree of risk of suffering any complication depends on the woman's health, the skill of the abortion practitioner, and how far advanced her pregnancy is. The risk of complications and death is also greater among Black woman and public-pay patients than among white women and private-pay patients, though researchers have been unable to determine why this is so.
The risk of death is widely touted to be less with abortion than with birth. This comparison is based on taking known abortion deaths -- detected mostly via statistical samples sent by state vital records offices to the National Center for Health Statistics -- and comparing them to maternal death rates from intensive maternal mortality studies. This means that we really don't know the risk of death from abortion.
But based on the scanty data available, the Centers for Disease Control has concluded that before 16 weeks, abortion is safer than childbirth, the risk is equal at sixteen weeks, and the risks of abortion death double every two weeks after week 16. This means that at 18 weeks, abortion is believed to be twice as risky as carrying to term, at 20 weeks four times as risky, and so forth. Most of the risk appears to be from the amount of effort it takes to dilate an unripe cervix to remove a large fetus, though injuries from bony fetal parts as the fetus is dismembered are also potentially catastrophic in later abortions.
Statistically, women who have undergone abortions also have a higher risk of suffering violent death -- suicide, homicide, or accident -- than women who have given birth or who have not been pregnant. The increased risk of suicide appears to be directly related to the abortion. The increased risk of homicide or accidental death has not been studied as much, and researchers currently can not determine if the increased risk is related to the abortion (strain on relationships, the known increased risk of drug or alcohol abuse in women who have undergone abortions), or if there is some common risk factor that leads women at higher risk of violent death to also be at higher risk of undergoing abortions.
I'd appreciate links to the most current research. From people who know what the heck they're taking about, not from people who are trusting the CDC and NARAL "fact sheets" which have been debunked to the point where anybody who hasn't relegated them to the status of bird-cage liners is delusional.