The infection rate in the UK has gone up and is close to the point where the virus starts spreading rapidly, government scientific advice says.
The so-called "R-number" is now between 0.7 and 1.0 - it needs to be kept below one in order to stay in control.
The rise in the figures is thought to be driven by the virus spreading in care homes and hospitals.
The effect of the changes to lockdown announced by the prime minister on Sunday is still unknown.
R is the number of people each infected person passes the virus on to, on average. If it is 2 then 10 infected people would pass it onto 20 others. But if is 0.5 then 10 infected people pass it onto 5 others.
The R-number had been sitting between 0.5 and 0.9. Any increase limits the ability of politicians to lift lockdown measures.
Health secretary Matt Hancock said the R-number was an "incredibly important figure" and it was important to note "we don't think that it is above one".
The number of new cases is still falling in the UK, but not as quickly as in the past. Sources said that progress was "getting closer to flat".
This could cause problems as low numbers of cases are needed to make it easier to perform mass testing and find the people anyone carrying the virus have come into contact with.
The latest analysis takes account of the spread of coronavirus in care homes, hospitals and more widely in society.
As the figures are based on patients ending up in hospital, they actually give a sense of the R-number from around three weeks ago.
That predates Boris Johnson's shift in England from "stay at home" to "stay alert", alongside encouraging some people back to work and allowing people to meet one person from outside their household outdoors.
The increase in the infection rate is said to be "consistent with" a significant fall in cases in the community and the epidemic and in turn the R-number being driven by care homes.
Sir Patrick Vallance, the government chief scientific adviser said: "To keep R below one and control the virus, it is vital that people stay alert and continue to follow the latest government guidelines to the letter."
Meanwhile, the Department of Health said it recorded another 384 deaths of people in the UK, bringing the total number to have died following a positive coronavirus test to 33,998.
What does this mean for me?
The big question is how the rise in the R number will affect plans to lift lockdown.
If R goes above 1 then there could be a "second peak" in cases and this is what politicians are trying to avoid.
The rise could mean the government has less wiggle room to lift restrictions and we'll have to live with them for longer.
But it is important to understand why the infection rate has increased.
One explanation is a large drop in cases in the general public means the R-number is now reflecting what is happening in care homes in hospitals.
That could mean there is still capacity to lift parts of lockdown in most parts of society.
Coronavirus disease modelling, published by the University of Cambridge, has also given a picture of how different regions of England are coping.
The figures do not perfectly match those from the SAGE group of government scientific advisors because it assesses multiple models to reach its conclusions.
The Cambridge study, backed by Public Health England, shows London has made the most progress with suppressing the virus, while it is proving far more stubborn in the north-east of England.
Those figures are more optimistic than other groups' calculations.
The Cambridge team calculated the figures by working backwards from the number of deaths. However, they have looked only at deaths confirmed with a genetic test rather than all deaths.
Similar work by the London School of Hygiene and Tropical Medicine puts the number for London at 0.6 and the South West at 0.9. They also showed the R-values were 0.8 in Wales, and 1 in both Scotland and Northern Ireland.
However, claims there are now just 24 cases a day in the capital and that it could soon be free of the virus have been criticised.
There were in fact 49 people admitted to London hospitals with Covid-19 yesterday and probably hundreds of cases that did not need hospital treatment.
"I am extremely worried about the media message that London could be coronavirus free in days," said Prof Matt Keeling, from the University of Warwick.
He added: "If people think London is coronavirus-free that could be dangerous, and could lead to complacency, undermining all the struggles and sacrifices that everyone has made so far. A relaxation of vigilance could easily see R increasing above 1, and a second epidemic wave."
Follow James on Twitter