One in three breast cancers detected by mammogram screening may actually be harmless, a study has suggested.
Data from five countries, including the UK, suggest some women may have had unnecessary treatment for cancers that were unlikely to kill them or spread.
As it is not possible to distinguish between lethal and harmless cancers, all are treated.
But advocates of screening insist it is a vital tool for early detection of cancerous cells.
Researchers from the Nordic Cochrane Centre in Denmark said their results showed cancer screening programmes could lead to "overdiagnosis".
Writing in the British Medical Journal (BMJ), they said: "Screening for cancer may lead to earlier detection of lethal cancers but also detects harmless ones that will not cause death or symptoms.
"The detection of such cancers, which would not have been identified clinically in someone's remaining lifetime, is called over-diagnosis and can only be harmful to those who experience it."
Professor Gilbert Welch, of the Dartmouth Institute for Health Policy, wrote an editorial on the finding in the BMJ.
He said: "It means that screening for cancer, in this case breast cancer, is a much closer call than has been previously advertised.
"It has the opportunity to help some women but it also has the consequence of leading others to be treated needlessly for cancer and that's not a trivial thing."
But Professor Julietta Patnick, director of NHS Cancer Screening Programmes, said one in eight women diagnosed with breast cancer through the NHS programme would have been missed without screening.
She said screening was estimated to save 1,400 lives a year in England alone.
Professor Patnick said: "Thanks to screening, one extra woman's life will be saved for every eight women diagnosed with breast cancer.
"By bringing forward the date of diagnosis (through early detection), we find those cancers that would otherwise not be caught until later in life by which time they could be fatal."
She also criticised the research, suggesting it made highly selective use of statistics, and ignored lifestyle changes which had increased breast cancer incidence, such as women waiting longer before having their first child.
The researchers looked at a range of statistics from five countries which had implemented screening programmes, including data for England and Wales from between 1971 and 1999.
The findings seem to confirm research published by the same team earlier this year.
Other recently compiled figures also show that UK death rates from breast, bowel, and male lung cancer are at their lowest since 1971.
The figures showing the fall in deaths from three of the most common cancers were compiled by Cancer Research UK, and are being put down to improved screening and better care.
'Don't be put off'
Dr Sarah Cant, from Breakthrough Breast Cancer, said she hoped the research on the incidence of harmless breast cancers would not discourage women from attending screening.
"Unfortunately, it is currently not possible to predict which cancers found through screening will develop aggressively and which will grow very slowly," she said.
"Based on all the current evidence, we believe the benefits of detecting breast cancer early still outweigh the risks."
She added that women needed to be given clear information about breast screening and it was important to remember that "while survival rates have increased greatly in recent years, just under 12,000 women still die from this disease each year in the UK."
Her view was echoed by Emma Pennery, of the charity Breast Cancer Care, who said: "Until it is possible to accurately determine the progression of cancers found through mammograms, screening remains an effective option for detecting breast cancers as soon as possible.
"As this review acknowledges this could lead to overtreatment in a percentage of cases.
"However, without screening women would face the prospect of having to wait for a visible symptom of cancer, such as a lump, to become apparent before treatment could start."
What do you think? Read some of your comments