An advert on Facebook for an app that provides a natural alternative to contraception has been banned by the UK's Advertising Standards Authority.
Claims that it was "highly accurate" and "provided a clinically tested alternative to other birth control methods" were found to be misleading.
The Swedish firm behind the Natural Cycles app was warned "not to exaggerate" its efficacy.
In response, the firm said it respected the outcome of the investigation.
It told the BBC that it removed the ad, which ran for approximately four weeks in mid-2017, as soon as it was notified of the complaint.
"We are committed to being open and transparent in our communications to ensure our message is clear and provides women with the information they need to determine if Natural Cycles is right for them. As part of these efforts, every advertisement undergoes a strict approval process," the firm said in a statement.
"Natural Cycles has been independently evaluated and cleared by regulators in Europe and the US based on clinical evidence demonstrating its effectiveness as a method of contraception."
The ASA said that the Facebook ad must not appear again in its current form.
Natural Cycles requires women to take their temperatures every day using a basal body thermometer and to enter the reading into the app, which also tracks a user's menstrual cycle.
The app uses an algorithm to determine a woman's daily fertility based on changes in basal body temperature.
Basal thermometers are able to detect a minor rise in temperature around the time of ovulation. Women will see a "use protection" warning appear on the app during their fertile days.
Launched in 2014, the app now has more than 300,000 users who pay a monthly or annual fee for the service.
It was invented by Swedish nuclear physicist Elina Berglund Scherwitzl and her husband.
It has previously been approved for use as a medical device by German inspection and certification organisation Tuv Sud, which means it can be used across the EU.
One former user - who asked to remain anonymous - told the BBC she had downloaded the app in September 2017 after being attracted by its Facebook campaign.
She said the ads had been "misleading" as they had suggested the software was simple to use. In practice, she explained, it had proved difficult to provide readings at matching times of the day since she had to care for two young children.
Nevertheless, she persevered but became pregnant.
When she contacted the firm to complain she said the staff had been "sympathetic but not apologetic" and had insisted she must have entered her data incorrectly.
Natural Cycles recently won approval from the US Food and Drug Administration, which described it as an effective method of contraception if "used carefully and correctly".
At the time, Terri Cornelison, assistant director for the health of women at the FDA's Center for Devices and Radiological Health, said: "Women should know that no form of contraception works perfectly, so an unplanned pregnancy could still result from correct usage of this device."
Reality Check: How do you know how effective your contraception is?
Contraception effectiveness is measured by looking at how often it fails - if 100 women use it over the course of a year, how many will become pregnant?
Looking at "typical use", which allows for some human error, Natural Cycle's founders say the figure would be seven women in every 100. With perfect use, they claim it is one in 100.
The most effective type of contraception is the implant, on which 0.0005 in every 100 women will become pregnant.
For the hormonal and copper coil it's 0.2 in 100 and just under one in 100 respectively.
About nine in every 100 women will become pregnant while relying on the pill.
But those figures for traditional contraception are taken from numerous studies done over many years. The two studies done so far on Natural Cycles only look at the app and mean their figures just are not comparable.
The Natural Cycles app relies on women taking their temperature at the same time every day and does not give an accurate fertility reading until three months of use.
The results can be skewed by various factors including:
- disturbed sleep or irregular sleep patterns
- drinking alcohol
- going to the toilet in the night
- taking painkillers in the night
- working shifts
In considering the complaint, which was lodged by three people, the ASA took expert advice and reviewed three published papers based on the accumulated data obtained from the app.
It found that there was a distinction between typical use of the app and the perfect-use scenario. It felt that the figures presented to users were based on the perfect user rather than typical user and, for the latter, the system could not be described as highly accurate.
It also found that presenting the statement "highly accurate" alongside the claim "clinically tested alternative to birth control methods" gave the impression "that the app was a precise and reliable method of preventing pregnancies which could be used in place of other established birth control methods".
Natural Cycles told the ASA that the claim "clinically tested alternative to birth control methods" was a quote from the news site Business Insider.
But, it added, the claims were backed by scientific evidence, including clinical trials.
Separately the app is being investigated by Sweden's Medical Products Agency (MPA) which told the BBC it has received approximately 60 complaints relating to unwanted pregnancies as a result of using the app.
It is due to publish its conclusions next week.