Trusts are wasting millions of pounds and losing thousands of service hours over-preparing for the European working time directive, figures shared exclusively with HSJ reveal.
Others risk prosecution and fines because the way they are assessing junior doctors' compliance with the directive does not match the EU measurement.
Data from 123 trusts shows the proportion of posts compliant with the directive using the EU's legal definition is up to twice as high as the Department of Health's figures.
NHS Employers head of programmes David Grantham said: "The situation may be better than people thought. Given the current drive for value for money you'd expect employers to be making every effort to make sure they're not going beyond what's necessary."
The directive, due to affect trainee doctors next August, will cut the maximum working week from 56 hours to 48 hours.
Instead of using the EU measurement, around half of trusts are thought to calculate their compliance with the directive using pay banding data for junior doctors on the new deal contract.
Those on bands 1a, 1b or 1c are deemed to be working within the 48 hours limit, while 2a and 2b posts are assumed to doing more hours than the directive permits. If all NHS bodies in the sample - around half of trusts - moved doctors down a band to ensure compliance under the new deal, they would unnecessarily lose 11,801 service hours.
The study, by doctor management software company Zircadian, said it estimated this meant around 248 more posts would be needed, costing roughly£12.4m.
And up to 26 per cent of band 1 posts are actually non-compliant with the directive, known as regulation four, according to the analysis.
Mr Grantham said the figures chimed with anecdotal evidence that trusts were underestimating how many doctors were compliant.
However, some trusts may be deliberately overcompensating to provide "extra headroom" when the regulation comes into force, he said.
Papworth Hospital foundation trust associate medical director of education and workforce Martin Goddard said squeezing doctors down to band 1 often had no cost benefits and lost them time for training.
Equally worrying was the fear of trusts risking prosecution by unwittingly breaking the law.
He said: "We've had no guidance centrally as to which method we should use. It has become a management nightmare."
The figures undermine statements made by Skills for Health, which is paid by the DH to help health service organisations achieve compliance. Skills for Health states on its website that discrepancies between the new deal and regulation four are "minor".
Zircadian found that in 54 trusts, compliance was up to twice as high as using the legal definition, 36 trusts showed no difference, and in 33 compliance was up to 40 per cent lower (see graph).
Mr Goddard, who is also a special adviser on the working time directive and new deal for NHS East of England, said he has highlighted the "fundamental differences" between the two calculations to the DH, but it was seen as "too difficult" to change.
The findings, showing around 60 per cent of posts are already meeting the 48-hour limit, also contrast with a Royal College of Surgeons survey last month warning fewer than half of trusts were compliant. A spokesperson called the Zircadian data a "sideshow" that must not distract managers on the front line.
Individual clinicians will be able to bring prosecutions against their employers using the EU law.
Zircadian managing director Henry Carleton warned failure to recognise the difference between the two measures could result in "clinical services and training time being unnecessarily trimmed and large amounts of money being wasted on phantom issues".
But a DH spokesman called this an "exaggeration". He said: "The department has always been clear that trusts have the responsibility to monitor working time directive compliance. The new deal working hours monitoring has only been used as a proxy for planning purposes."
It will be publishing guidance on the issue.
Main differences between new deal and EU measurement of compliance with the working time directive:
Average working hours under the new deal are normally worked out over an eight week period and take into account "prospective cover" for doctors who work rotas normally carried out by colleagues away on leave.
The European measurement does not include prospective cover and uses a 26-week reference period for doctors in training.