In raw figures, their first round of national test results showed that more than half the 11-year-olds achieved below level 4 - the target for their age group - in English and maths. Seven-year-olds did much better, with about 80 per cent reaching national standards, set at level 2.
The picture at the age of 14 is more confusing. The Government has suggested that results at lower secondary are better than in the juniors, although not much, with more than half at level 5 or above. But another reading of the figures shows that they have actually done quite a bit worse.
The Government says that an average 14-year-old should reach level 5 or 6. This suggests that half the pupils would be expected to be at level 5 and below, while the other half should be at level 6 or above.
However, the figures - not the ones emphasised by the Government - show that in the KS3 tests, the percentages getting level 6 and above were 20 per cent in English, 33 in maths and 25 per cent in science.
The figures at all ages have been criticised on many fronts. Teacher unions and other organisations have said the tests were flawed, the curriculum was "grossly overcrowded" and the schools were under-resourced.
While there was a great deal invested in training for Year 2 teachers when the tests at seven began in 1991, there has been no equivalent investment at key stage 2.
The Government has now pledged Pounds 3 million for such training, and the slimmed-down version of the national curriculum itself took effect several months after these tests.
The National Association for Primary Education says curriculum overcrowding "led to less emphasis upon the teaching of English and mathematics in the crucially important early years in school".
They argue that at KS2 "the need for a re-examination of the test levels is clearly demonstrated by the imbalance between the results achieved in science (70 per cent reaching or exceeding an average standard) compared with mathematics (44 per cent)".
There is also the simple point that in the first round of testing, there are bound to be teething problems, both within the tests themselves and in schools.
The English tests, and particularly the external marking, were heavily criticised at both key stages 2 and 3.
And at both these stages, teachers' own assessments in all subjects (apart from KS2 science) were much higher than the test results, although they were still not that good (only 56 per cent level 4 and above in KS2 English).
At key stage 1, results have improved over the years, particularly in the numbers doing better than the average, and reaching level 3.
At Year 6, it was widely held that children did not have enough time to do the maths tests, and an extra 15 minutes will be added in this year's round.
David Winkley, head of Birmingham Grove Junior School, Birmingham, which has been much in the news this week for its accelerated maths programmes, is convinced that the KS2 maths test was seriously flawed.
He points to four of last year's Year 6 pupils who attained grade C in the maths GCSE (equivalent to level 7 or 8), but only achieved level 5 in the KS2 test. Two of these children also sat the entrance exam to a top independent school, King Edward's, and came out in the top group in a test taken by children from every prep school in the region.
"I think those SATs were very unreliable indeed, " says Dr Winkley. "They had an absurd time limit."
Launching the results last week, the Education and Employment Secretary, Gillian Shephard, accepted that there were all sorts of provisos about the results, but added: "We have to accept that standards of attainment are not good enough in English and maths."
Meanwhile, other questions are being raised about the national curriculum levels. These were pitched by the Task Group on Assessment and Testing in 1987, based on their best judgments, and some critics argue that they may not be right. If half of 11-year-olds have not reached level 4, are the children doing badly or is the level wrong?
Chris Jones, assistant secretary for curriculum 5-14 at the School Curriculum and Assessment Authority, believed the levels were pitched right.
The Dearing Review had provided opportunities for consultation and review, and there had been monitoring over five years. SCAA felt confident it had been able to iron out any difficulties with the levels, he said.