Some of the schools with the most experience of the literacy hour have seen big falls in their English key stage 2 test results when compared with last year.
The Government is pinning its hopes of 11-year-olds achieving stringent targets in 2002 on the success of its literacy and numeracy strategies.
National Literacy Project schools were the first to pilot the literacy hour, from 19967, and as a group have improved faster than other schools.
However, a TES analysis of the 1998 results for individual schools which piloted literacy hours reveals a mixed performance.
In five authorities - Bristol, Lambeth, Liverpool, Tower Hamlets, and Waltham Forest - the pilot schools improved faster than others in their authorities, and significantly above national rates.
Lambeth's first nine pilot schools were the top scorers, with a 10.8 percentage-point improvement on their 1997 results on average.
The south London authority as a whole saw an improvement of 4.5 percentage points, well above the England average of 1.6 points. In both 1997 and 1998, its first pilot schools were performing above Lambeth averages.
But in another six areas - Islington, Manchester, Newcastle, Newham, Sandwell, and Southwark - the results for the pilot schools were down on 1997, despite improvements overall in the authorities.
In Newcastle, results at its first 16 literacy-hour schools were down 2.1 percentage points, to 35.1 per cent of pupils achieving the standard expected for their age. Yet, English scores for the authority as a whole were up 3.8 points to 54.8 per cent.
Eighteen authorities took part in the first National Literacy Project pilots, which targeted schools with the poorest reading and writing results and biggest problems, including high levels of deprivation and pupils with English as an additional language.
The authorities say that other evidence - inspection reports, key stage 1 results, and standardised tests - reveals the literacy strategy is raising standards.
A spokeswoman for Manchester said a fairer comparison is between results in 1996 and 1998 - when its pilot-school results improved from 33.2 to 42.6 per cent of pupils achieving the standard expected for their age. However, there were objections to the publication of the 1996 results on the grounds that the key-stage tests were not yet "bedded in".
Both David Bell, Newcastle's director of education, and Jan Roman, Sandwell's senior primary adviser, point to big improvements in test results for seven-year-olds, as indicating the literacy project will deliver on government-set targets at age 11.
Mrs Roman said: "We believe the literacy project is going to make a significant difference and is already impacting - but not yet on KS2 statistics, which is unfortunately what it is being measured by. We need some more time."
The Government is expecting a more consistent pattern of improvement over time, said a spokesperson, and less of the extreme variation in The TES analysis.