A Dutch inquiry into the UN's climate science panel has found "no errors that would undermine the main conclusions" on probable impacts of climate change.
However, it says the Intergovernmental Panel on Climate Change (IPCC) should be more transparent in its workings.
The Dutch parliament asked for the inquiry after two mistakes were identified in the IPCC's 2007 report.
The inquiry is the latest in a series that have largely backed "mainstream" climate science against detractors.
The Netherlands Environmental Assessment Agency (PBL) does not give the panel a completely clean bill of health, however.
Whereas the IPCC's landmark Fourth Assessment (AR4) from 2007 "conclusively shows" that impacts of human-induced climate change are already tangible in many places around the world and will become more serious as temperatures increase, PBL also says the foundation for some of the specific projections "could have been made more transparent".
The Netherlands inquiry adds that the IPCC's summaries tended to emphasise "worst-case scenarios".
However, this was disputed by scientists who had played a leading role in AR4.
"The net impacts of climate change are not beneficial," said David Vaughan, science leader at the British Antarctic Survey, who co-ordinated the AR4 chapter on polar impacts.
Martin Parry, visiting professor at the Grantham Institute for Climate Change Research at Imperial College London who co-chaired AR4 Working Group 2 on climate impacts, welcomed the PBL report.
"We welcome the conclusion of this report, which is esentially that our conclusions are safe, sound and reliable," he said.
"The IPCC is about to venture into the next assessment; so it's important that we learn from these issues, and it's important not to be defensive, and I think that's how the IPCC is approaching things now."
PBL's central remit was to look at how accurately the IPCC's chapters on regional projections of climate impacts reflected the science available, and whether those chapters were adequately distilled into the summary that was given to government representatives for signing off.
The panel has already admitted making a mistake over the projected date for disappearance of glaciers in the Himalayas.
It projected that glaciers could largely disappear by 2035, which the IPCC acknowledges (and the PBL inquiry confirms) is highly unlikely.
A furore erupted in the Netherlands over a different claim in AR4, namely that 55% of the country was vulnerable to flooding because it was situated below sea level.
The source of the claim was a report from PBL itself, which said that 55% of the Netherlands was prone to flooding.
But that report said only 26% of the country was at risk because it lies below sea level, with the remainder affected by river flooding.
PBL now accepts the blame for the mistake lies within its own doors.
"We acknowledge that this error was not the fault of the IPCC... the error was made by a contributing author from the PBL, and the (co-ordinating) lead authors (of AR4 chapters) are not to blame for relying on Dutch information provided by a Dutch agency," it said.
PBL also says it has uncovered another minor error in AR4's summary.
The IPCC said that by the year 2020, between 75 million and 250 million Africans would be at risk of "water stress" (ie not having enough water). PBL says that based on the science available, the figures should be 90-220 million - but that the IPCC projections fit within the "range of uncertainty" in the science.
However, Nigel Arnell, head of the Walker Centre at the University of Reading who led the water chapter in AR4, disputed the PBL assessment.
"The figures are based on a series of assumptions and calculations," he said.
"I think the way in which it was projected with a wide range encapsulated the huge uncertainties, and we think that (narrowing it to) 90-220 million is an over-interpretation of the information that the chapter authors had at the time."
Overall, there were 35 instances where PBL cited errors or made other comments. The vast majority were minor, such as typographical errors or the mis-labelling of a graph; Professor Parry and other IPCC authors accept 12, and have noted them as errata on the IPCC website.
PBL closes its investigation with a number of recommendations for the IPCC, including setting up a public website for the submission of information on possible errors, additional funding for staff to assist with quality control, and taking care with public statements that "could be perceived... as heightening the projected impacts of climate change".
This is the latest in a series of reviews and inquiries that have endorsed the central conclusions of mainstream climate science, while finding issues of concern in how it is practiced.
Two reviews into issues arising from the emails hacked from the University of East Anglia's Climatic Research Unit (CRU) late last year have concluded there was no evidence of scientific malpractice or manipulation of data.
However, they asked for greater openness, compliance with Freedom of Information law and better collaboration with professional statisticians.
The third review into the CRU issue will be published on Wednesday.
Meanwhile, the Sunday Times was recently forced to apologise for claiming that IPCC projections on die-back of the Amazon rainforest were unsubstantiated.
The main international review of the IPCC - conducted by the InterAcademy Council, a network of national science academies such as the UK's Royal Society - is ongoing, with formal publication due in October.