As recently as the 1950s, what came out of the delivery room was pretty much a surprise; the standard prenatal exam involved placing a stethoscope on the mother’s stomach and listening for signs of life. That changed in the mid-1960s, when doctors started extracting fluid from the womb and checking for abnormalities. With that test, known as amniocentesis, prospective parents could spot some congenital defects by the 20th we of pregnancy, when abortion was still possible. Later, a technique called chorionic villus sampling (CVS) produced the same information at 10 weeks. But for families prone to hereditary disorders, neither method was ideal. If the test revealed a defect, parents had to choose between aborting their child or living with the consequences.

Today some high-risk families are avoiding the dilemma altogether. Through a new procedure called BABI (blastomere analysis before implantation), a couple can conceive several embryos in test tubes and discard those exhibiting known defects. The first step is fairly routine; doctors give the woman a drug to stimulate ovulation, then extract eggs from her ovaries and mix them with her partner’s sperm. Instead of returning embryos directly to the womb, the doctors let them grow in a culture until they consist of eight cells. Since each cell contains the individual’s entire genetic code, the doctors pull a cell from each embryo to see whether it harbors a mutation for the disease in question. Defective embryos aren’t Placed in the womb, but those that are implanted can still develop normally. Dr. Mark Hughes, the Houston geneticist who pioneered BABI, says it has yielded at least a dozen healthy pregnancies in the past two years.

So far, the procedure has been used only to test embryos for such clear-cut hereditary defects as TaySachs disease and Duchenne muscular dystrophy. But as researchers identify genes that help predispose people to cancer or high blood pressure, prospective parents could become choosier about which embryos to keep. Indeed, some critics worry that parents could use a procedure like BABI to select embryos according to their likely stature, disposition or intelligence. In his book “The Human Body Shop,” Andrew Kimbrell of the anti-biotech Foundation on Economic Trends declares with alarm that “advances in the preimplantation genetic screening of embryos have initiated the journey toward the ‘perfect baby’.” whether that’s such a bad thing is debatable; some analysts see little harm in giving one’s children every perceived advantage. But good or bad, made-to-order babies are still a pipe dream.

The reason is that interesting traits such as intelligence or personality don’t stem from single genes. The 23 chromosome pairs we carry in the nucleus of each cell contain 3 billion units of DNA, arranged in linear sequence. Genes are simply functional groups of DNA units that specify the makeup of proteins. Those proteins interact with each other and their surroundings to construct and operate the body. Of the 100,000 or so genes contained in our chromosomes, half may be involved in the workings of the brain. If scientists knew exactly where they were located and how their interactions affected intelligence, spotting a smart embryo would be a realistic goal. But as Hughes observes, “We don’t know which genes are relevant [or] what controls them.” The ongoing Human Genome Project will eventually produce the raw data needed to study them, by spelling out the genetic sequence one unit at a time. But it won’t produce a simple probe for intelligence.

Nor will it enable us to design superior beings. “If one wanted to breed people like racehorses, the technology for that has been with us for hundreds of years,” says geneticist Phil Leder of Harvard and the Howard Hughes Medical Institute. “Genetic techniques offer very little that is new.” Scientists are optimistic about learning to treat single-gene diseases. But there’s a big difference between giving a person therapeutic doses of a missing gene and reinventing the embryo. As geneticist Eric Lander of MIT puts it, “We can tinker at the edges, but we are talking about an understanding of single components in a system that has 100,000 components.” And as long as that’s the case, the best way to produce smarter kids will be to educate them.