The number of healthy babies lost as a result of Down's syndrome testing is "unjustifiable", a charity has claimed.
Down Syndrome Education International estimates two healthy foetuses are miscarried for every three found to have the condition.
Screening is offered to every woman, and more invasive tests - which can cause miscarriage - to those found to be "high risk".
But the research was condemned as 'unhelpful' by another Down's charity.
It's not as clean and effective as everyone likes to make out - there should be much more thorough debate about these tests
Down Syndrome Education International
Down's syndrome is a genetic condition caused by a baby having an extra chromosome.
Babies with the condition often have learning difficulties and can have serious heart problems, although many people with Down's have a long lifespan and can live semi-independently.
For these reasons, the charity, which supports parents raising children with Down's syndrome, is firmly against the idea of genetic screening, and the subsequent abortion of foetuses which test positive.
Its own analysis suggests that, in an average year in the UK, screening confirms 660 foetuses with Down's. Most women given these test results opt to abort the pregnancy.
Most women are initially offered either a blood test or an ultrasound test to look for signs of Down's in the foetus.
A positive test at this point does not mean Down's is highly likely - in some cases, only one in 20 women who screen positive actually have an affected baby.
To confirm or rule it out, a further test is offered - either amniocentesis and chorionic villus sampling, both of which use a fine needle to take tissue or fluid from inside the womb.
However, the charity suggests that the risks linked to these invasive tests, and the high number carried out, lead to 400 miscarriages in pregnancies unaffected by Down's.
Some research suggests that the miscarriage rate following chorionic villus sampling may be as high as 2%, the researchers said, which, if true, would raise that toll closer to 600.
National guidelines now recommend a newer version of screening - the "combined test", which may be more accurate in predicting which women are at higher risk of carrying a baby with Down's.
However, the researchers said that despite a lower number of invasive tests, the heavier reliance of this method on chorionic villus sampling earlier in pregnancy could actually mean an increase in the number of healthy babies lost if the miscarriage rate using this method was closer to 2%.
Frank Buckley, the charity's chief executive, said that many women misunderstood the risks involved, and were not necessarily making an informed choice: "It's not as clean and effective as everyone likes to make out - there should be much more thorough debate about these tests."
'Unfortunate and unhelpful
However, the Down's Syndrome Association, while saying it was "saddened" by the loss of any unaffected pregnancy, was critical of the research.
Professor Jennifer Wishart, said: "The 'two for three' approach taken in the DownsEd research is unfortunate and unhelpful to ongoing debates about screening.
"There has never been any dispute that current diagnostic techniques have associated risk and that women need to be better and fully informed of these risks.
"The 'two for three' emphasis - that two healthy babies are lost for every three babies with Down's syndrome prenatally detected - simply muddies the waters."
The Department of Health responded that the aim of ante-natal screening was to offer women the information they needed to make the right choice.
"The screening programme supports consistent training and information for staff - to ensure that screening is both effective and safe for all women, a quality assurance system has been set up."